Datasets:

ppxscal

/

academic_embeddings_cosimrank_test_bfs

like 0

A File System Based on Concept Analysis We present the design of a file system whose organization is based on Concept Analysis "á la Wille-Ganter". The aim is to combine querying and navigation facilities in one formalism. The file system is supposed to offer a standard interface but the interpretation of common notions like directories is new. The contents of a file system is interpreted as a Formal Context, directories as Formal Concepts, and the sub-directory relation as Formal Concepts inclusion. We present an organization that allows for an efficient implementation of such a Conceptual File System.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

Parameterized complexity for the database theorist

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

A Weekly Supervised Speech Enhancement Strategy using Cycle-GAN Nowadays, due to the application of deep neural network (DNNS), speech enhancement (SE) technology has been significantly developed. However, most of current approaches need the parallel corpus that consists of noisy signals, corresponding speech signals and noise on the DNNs training stage. This means that a large number of realistic noisy speech signals is difficult to train the DNNs. As a result, the performance of the DNNs is restricted. In this research, a new weakly supervised speech enhancement approach is proposed to break this restriction, using the cycle-consistent generative adversarial network (CycleGAN). There are two stage for our methods. In training stage, a forward generator is employed to estimate ideal time-frequency (T-F) mask and an inverse generator is utilized to acquire noisy speech magnitude spectrum (MS). Additionally, two discriminators are used to distinguish the real clean and noisy speech from generated speech, respectively. In enhancement stage, the T-F mask is directly estimated by using the well-trained forward generator for speech enhancement. Experimental results indicate that our strategy can not only achieve satisfied performance for non-parallel data, but also acquire the higher score in speech quality and intelligibility for the DNN-based speech enhancement using parallel data.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Learning Topic Representation For Smt With Neural Networks Statistical Machine Translation (SMT) usually utilizes contextual information to disambiguate translation candidates. However, it is often limited to contexts within sentence boundaries, hence broader topical information cannot be leveraged. In this paper, we propose a novel approach to learning topic representation for parallel data using a neural network architecture, where abundant topical contexts are embedded via topic relevant monolingual data. By associating each translation rule with the topic representation, topic relevant rules are selected according to the distributional similarity with the source text during SMT decoding. Experimental results show that our method significantly improves translation accuracy in the NIST Chinese-to-English translation task compared to a state-of-the-art baseline.

A Policy Description Language A policy describes principles or strategies for a plan of action designed to achieve a particular set of goals. We define a policy as a function that maps a series of events into a set of actions. In this paper we introduce PDL, a simple but expressive language to specify policies. The design of the language has been strongly influenced by the action languages of Geffner and Bonet (Geffner

Reasoning about Policies using Logic Programs We use a simplied version of the Policy DescriptionLanguage PDL introduced in (Lobo, Bhatia, & Naqvi1999) to represent and reason about policies. In PDLa policy description is a collection of Event-ConditionAction-Rules that denes a mapping from event historiesinto action histories. In this paper we introduce thegeneration problem: nding an event history generatingan action history, and state its complexity. Becauseof its high complexity we present a logic programmingbased...

On logical foundations of active databases In this chapter, we present work on logical foundations of active databases. After introducing the basic notions and terminology, we give a short overview of research on foundations of active rules. Subsequently, we present a specific state-oriented logical approach to active rules which aims at combining the declarative semantics of deductive rules with the possibility to define updates in the style of production rules. The resulting language Statelog models (flat) transactions as a sequence of intermediate transitions, where each transition is defined using deductive rules. Since Statelog programs correspond to a specific class of locally stratified logic programs, they have a unique intended model. Finally, after studying further fundamental properties like expressive power and termination behavior, a Statelog framework for active rules is presented. Although the framework is surprisingly simple, it allows to model many essential features of active rules, including immediate and deferred rule execution, and composite events. Different alternatives for enforcing termination are proposed leading to tractable subclasses of the language. Finally, we show that certain classes of Statelog programs correspond to Datalog programs with production rule semantics (i.e., with inflationary or noninflationary fixpoint semantics).

Well founded semantics for logic programs with explicit negation . The aim of this paper is to provide asemantics for general logic programs (with negation bydefault) extended with explicit negation, subsumingwell founded semantics [22].The Well Founded semantics for extended logicprograms (WFSX) is expressible by a default theorysemantics we have devised [11]. This relationshipimproves the cross--fertilization between logic programsand default theories, since we generalize previousresults concerning their relationship [3, 4, 7, 1, 2],and there is...

Representing actions in logic programs and default theories a situation calculus approach We address the problem of representing common sense knowledge about action domains in the formalisms of logic programming and default logic. We employ a methodology proposed by Gelfond and Lifschitz which involves first defining a high-level language for representing knowledge about actions, and then specifying a translation from the high-level action language into a general-purpose formalism, such as logic programming. Accordingly, we define a high-level action languageAE, and specify sound and complete translations of portions ofAEinto logic programming and default logic. The languageAEincludes propositions that represent “static causal laws” of the following kind: a fluent formula ψ can be made true by making a fluent formula true (or, more precisely, ψ is caused whenever is caused). Such propositions are more expressive than the state constraints traditionally used to represent background knowledge. Our translations ofAEdomain descriptions into logic programming and default logic are simple, in part because the noncontrapositive nature of causal laws is easily reflected in such rule-based formalisms.

Representing action and change by logic programs We represent properties of actions in a logic programming language that uses both classical negation and negation as failure. The method is applicable to temporal projection problems with incomplete information, as well as to reasoning about the past. It is proved to be sound relative to a semantics of action based on states and transition functions.

The well-founded semantics for general logic programs A general logic program (abbreviated to “program” hereafter) is a set of roles that have both positive and negative subgoals. It is common to view a deductive database as a general logic program consisting of rules (IDB) slttmg above elementary relations (EDB, facts). It is desirable to associate one Herbrand model with a program and think of that model as the “meaning of the program, ” or Its“declarative semantics. ” Ideally, queries directed to the program would be answered in accordance with this model. Recent research indicates that some programs do not have a “satisfactory” total model; for such programs, the question of an appropriate partial model arises. Unfounded sets and well-founded partial models are introduced and the well-founded semantics of a program are defined to be its well-founded partial model. If the well-founded partial model is m fact a total model. it is called the well-founded model. It n shown that the class of programs possessing a total well-founded model properly includes previously studied classes of “stratified” and “locally stratified” programs,The method in this paper is also compared with other proposals in the literature, including Clark’s“program completion, ” Fitting’s and Kunen’s 3-vahred interpretations of it, and the “stable models”of Gelfond and Lifschitz.

Restricted Monotonicity A knowledge representation problem can be sometimesviewed as an element of a family of problems,with parameters corresponding to possibleassumptions about the domain under consideration.When additional assumptions are made,the class of domains that are being described becomessmaller, so that the class of conclusions thatare true in all the domains becomes larger. Asa result, a satisfactory solution to a parametricknowledge representation problem on the basis ofsome nonmonotonic...

Efficient top-down computation of queries under the well-founded semantics The well-founded model provides a natural and robust semantics for logic programs with negative literals in rule bodies. Although various procedural semantics have been proposed for query evaluation under the well-founded semantics, the practical issues of implementation for effective and efficient computation of queries have been rarely discussed.

Improving the I/O Performance of Real-Time Database Systems with Multiple-Disk Storage Structures

Input space versus feature space in kernel-based methods. This paper collects some ideas targeted at advancing our understanding of the feature spaces associated with support vector (SV) kernel functions. We first discuss the geometry of feature space. In particular, we review what is known about the shape of the image of input space under the feature space map, and how this influences the capacity of SV methods. Following this, we describe how the metric governing the intrinsic geometry of the mapped surface can be computed in terms of the kernel, using the example of the class of inhomogeneous polynomial kernels, which are often used in SV pattern recognition. We then discuss the connection between feature space and input space by dealing with the question of how one can, given some vector in feature space, find a preimage (exact or approximate) in input space. We describe algorithms to tackle this issue, and show their utility in two applications of kernel methods. First, we use it to reduce the computational complexity of SV decision functions; second, we combine it with the Kernel PCA algorithm, thereby constructing a nonlinear statistical denoising technique which is shown to perform well on real-world data.

Near-Optimal Parallel Prefetching and Caching Recently there has been a great deal of interest in the operating systems research community in prefetching and caching data from parallel disks, as a technique for enabling serial applications to improve input--output (I/O) performance. In this paper, algorithms are considered for integrated prefetching and caching in a model with a fixed-size cache and any number of backing storage devices (disks). The integration of caching and prefetching with a single disk was previously considered by Cao, Felten, Karlin, and Li. Here, it is shown that the natural extension of their aggressive algorithm to the parallel disk case is suboptimal by a factor near the number of disks in the worst case. The main result is a new algorithm, reverse aggressive, with near-optimal performance for integrated prefetching and caching in the presence of multiple disks.

Scheduling parallel I/O operations The I/O bottleneck in parallel computer systems has recently begun receiving increasing interest. Most attention has focused on improving the performance of I/O devices using fairly low-level parallelism in techniques such as disk striping and interleaving. Widely applicable solutions, however, will require an integrated approach which addresses the problem at multiple system levels, including applications, systems software, and architecture. We propose that within the context of such an integrated approach, scheduling parallel I/O operations will become increasingly attractive and can potentially provide substantial performance benefits.We describe a simple I/O scheduling problem and present approximate algorithms for its solution. The costs of using these algorithms in terms of execution time, and the benefits in terms of reduced time to complete a batch of I/O operations, are compared with the situations in which no scheduling is used, and in which an optimal scheduling algorithm is used. The comparison is performed both theoretically and experimentally. We have found that, in exchange for a small execution time overhead, the approximate scheduling algorithms can provide substantial improvements in I/O completion times.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

A deep-learning based feature hybrid framework for spatiotemporal saliency detection inside videos. Although research on detection of saliency and visual attention has been active over recent years, most of the existing work focuses on still image rather than video based saliency. In this paper, a deep learning based hybrid spatiotemporal saliency feature extraction framework is proposed for saliency detection from video footages. The deep learning model is used for the extraction of high-level features from raw video data, and they are then integrated with other high-level features. The deep learning network has been found extremely effective for extracting hidden features than that of conventional handcrafted methodology. The effectiveness for using hybrid high-level features for saliency detection in video is demonstrated in this work. Rather than using only one static image, the proposed deep learning model take several consecutive frames as input and both the spatial and temporal characteristics are considered when computing saliency maps. The efficacy of the proposed hybrid feature framework is evaluated by five databases with human gaze complex scenes. Experimental results show that the proposed model outperforms five other state-of-the-art video saliency detection approaches. In addition, the proposed framework is found useful for other video content based applications such as video highlights. As a result, a large movie clip dataset together with labeled video highlights is generated.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Learning Topic Representation For Smt With Neural Networks Statistical Machine Translation (SMT) usually utilizes contextual information to disambiguate translation candidates. However, it is often limited to contexts within sentence boundaries, hence broader topical information cannot be leveraged. In this paper, we propose a novel approach to learning topic representation for parallel data using a neural network architecture, where abundant topical contexts are embedded via topic relevant monolingual data. By associating each translation rule with the topic representation, topic relevant rules are selected according to the distributional similarity with the source text during SMT decoding. Experimental results show that our method significantly improves translation accuracy in the NIST Chinese-to-English translation task compared to a state-of-the-art baseline.

MaSSIVE: the Mass Storage System IV Enterprise The Mass Storage System IV Enterprise (MaSSIVE) a fourth-generation mass storage system that provides a file system service to teraflops computing systems, is described. A bitfile managed by MaSSIVE can be a complete, self-contained Unix file system specific to a particular MaSSIVE client. MaSSIVE will stage whole file systems from archival storage onto online storage devices and then provide its clients with raw block access to the staged file systems. It will comply with the IEEE Mass Storage System Reference Model, and will consist of cooperating processes distributed across a network of supermicrocomputers. The client interface to a file system will be implemented through a custom device driver that will permit I/O to MaSSIVE online storage drives. The device interface through a high-bandwidth data fabric to the storage device will use common device-controller-specific protocols. As file systems are unmounted by clients or clients disconnect from the data fabric, file systems can be migrated back to archive storage and removed from online storage

File servers for network-based distributed systems A file server provides remote centralized storage of data to workstations connected to it via a communicatmn network; it facilitates data sharing among autonomous workstations and support of inexpensive workstations that have limited or no secondary storage. Various characteristics of file servers and the corresponding implementation issues based on a survey of a number of experimental file servers are discussed and evaluated in this paper. Particular emphasis is placed on the problem of atomm update of data stored in a file server. The design issues related to the scope of atomic transactions and the granularity of data access supported by a file server are studied in detail.

Storage systems for national information assets An industry-led collaborative project, called the National Storage Laboratory (NSL), has been organized to investigate technology for storage systems that will be the future repositories for the national information assets. Lawrence Livermore National Laboratory through its National Energy Research Supercomputer Center (NERSC) is the operational site and the provider of applications. It is anticipated that the integrated testbed system will represent a significant advance in the technology for distributed storage systems capable of handling gigabyte class files at gigabit-per-second data rates. The NSL collaboration is undertaking research in four areas: network-attached storage; multiple, dynamic, distributed storage hierarchies; layered access to storage system services; and storage system management. An overview of the prototype storage system is given. Three application domains have been chosen to test and demonstrate the system's effect on scientific productivity; climatic models, magnetic fusion energy models, and digital imaging

I/O For Tflops Supercomputers Scalable parallel computers with TFLOPS (Trillion FLoating Point Operations Per Second) performance levels are now under construction. While we believe TFLOPS processor technology is sound, we believe the software and I/O systems surrounding them need improvement. This paper describes our view of a proper system that we built for the nCUBE parallel computer and which is now commercially available. The distinguishing feature of our system is that scalable parallelism is implicit rather than explicit. We did not base our system on new commands, system calls, or languages. Instead, we extended some aspects of Unix® to add parallelism while keeping these aspects unchanged for nonparallel programs. The result is a system that lets one use a future TFLOPS parallel computer without knowing parallel programming. As parallel versions of standard compilers arrive, and large data sets get distributed over multiple I/O devices, then standard Unix commands will run arbitrary mixtures of parallel and nonparallel programs and I/O devices. One gets scalable computing and I/O rates whenever a command includes only parallel components.

The High Performance Storage System The National Storage Laboratory (NSL) was organized to develop, demonstrate and commercialize technology for the storage systems that will be the future repositories for the national information assets. Within the NSL four Department of Energy laboratories and IBM Federal Systems Company pooled their resources to develop an entirely new High Performance Storage System (HPSS). The HPSS project concentrates on scalable parallel storage systems for highly parallel computers as well as traditional supercomputers and workstation clusters. Concentrating on meeting the high end of storage system and data management requirements, HPSS is designed using network-connected storage devices to transfer data at rates of 100 million bytes per second and beyond. The resulting products will be portable to many vendor's platforms. The three year project is targeted to be complete in 1995. This paper provides an overview of the requirements, design issues, and architecture of HPSS, as well as a description of the distributed, multi-organization industry and national laboratory HPSS project.

Analysis of striping techniques in robotic storage libraries In recent years advances in computational speed have been the main focus of research and development in high performance computing. In comparison, the improvement in I/O performance has been modest. Faster processing speeds have created a need for faster I/O as well as for the storage and retrieval of vast amounts of data. The technology needed to develop these mass storage systems exists today. Robotic storage libraries are vital components of such systems. However, they normally exhibit high latency and long transmission times. We analyze the performance of robotic storage libraries and study striping as a technique for improving response time. Although striping has been extensively studied in the content of disk arrays, the architectural differences between robotic storage libraries and arrays of disks suggest that a separate study of striping techniques in such libraries would be beneficial.

Extended ephemeral logging: log storage management for applications with long lived transactions Extended ephemeral logging (XEL) is a new technique for managing a log of database activity subject to the general assumption that the lifetimes of an application’s transactions may be statistically distributed over a wide range. The log resides on nonvolatile disk storage and provides fault tolerance to system failures (in which the contents of volatile main memory storage may be lost). XEL segments a log into a chain of fixed-size FIFO queues and performs generational garbage collection on records in the log. Log records that are no longer necessary for recovery purposes are “thrown away” when they reach the head of a queue; only records that are still needed for recovery are forwarded from the head of one queue to the tail of the next. XEL does not require checkpoints, permits fast recovery after a crash and is well suited for applications that have a wide distribution of transaction lifetimes. Quantitative evaluation of XEL via simulation indicates that it can significantly reduce the disk space required for the log, at the expense of slightly higher bandwidth for log information and more main memory; the reduced size of the log permits much faster recovery after a crash as well as cost savings. XEL can significantly reduce both the disk space and the disk bandwidth required for log information in a system that has been augmented with a nonvolatile region of main memory.

A file system for continuous media The Continuous Media File System, CMFS, supports real-time storage and retrieval of continuous media data (digital audio and video) on disk. CMFS clients read or write files in “sessions,” each with a guaranteed minimum data rate. Multiple sessions, perhaps with different rates, and non-real-time access can proceed concurrently. CMFS addresses several interrelated design issues; real-time semantics fo sessions, disk layout, an acceptance test for new sessions, and disk scheduling policy. We use simulation to compare different design choices.

Multi-disk B-trees

Random duplicated assignment: an alternative to striping in video servers An approach is presented for storing video data in large disk arrays. Video data is stored by assigning a number of copies of each data block to different, randomly chosen disks, where the number of copies may depend on the popularity of the corresponding video data. The approach offers an interesting alternative to the well-known striping techniques. Its use results in smaller response times and lower disk and RAM costs if many continuous variable-rate data streams have to be sustained simultaneously. It also offers some practical advantages relating to reliability and extendability.Based on this storage approach, three retrieval algorithms are presented that determine, for a given batch of data blocks, from which disk each of the data blocks should be retrieved. The performance of these algorithms is evaluated from an average-case as well as a worst-case perspective.

Using Deep Belief Nets to Learn Covariance Kernels for Gaussian Processes We show how to use unlabeled data and a deep belief net (DBN) to learn a good covariance kernel for a Gaussian process. We first learn a dee p generative model of the unlabeled data using the fast, greedy algorithm intro duced by (7). If the data is high-dimensional and highly-structured, a Gaussian kernel applied to the top layer of features in the DBN works much better than a similar kernel applied to the raw input. Performance at both regression and classifi cation can then be further improved by using backpropagation through the DBN to discriminatively fine-tune the covariance kernel. a mixture P yn p(xn|yn)p(yn) and then infer p(yn|xn), (15) attempts to learn covariance kernels based on p(X), and (10) assumes that the decision boundaries should occur in regions where the data density, p(X), is low. When faced with high-dimensional, highly-structured data, however, none of the existing approaches have proved to be particularly successful. In this paper we exploit two properties of DBN's. First, they can be learned efficiently from unla- beled data and the top-level features generally capture sig nificant, high-order correlations in the data. Second, they can be discriminatively fine-tuned using backp ropagation. We first learn a DBN model of p(X) in an entirely unsupervised way using the fast, greedy learning algorithm introduced by (7) and further investigated in (2, 14, 6). We then use this gener ative model to initialize a multi-layer, non-linear mapping F(x|W), parameterized by W , with F : X ! Z mapping the input vectors in X into a feature space Z. Typically the mapping F(x|W) will contain millions of parameters. The top-level features produced by this mapping allow fairly accurate reconstruction of the input, so they must contain most of the information in the input vector, but they express this information in a way that makes explicit a lot of the higher-order structure in th e input data.

On sets with efficient implicit membership tests This paper completely characterizes the complexity of implicit membership testing in terms of the well-known complexity class OptP, optimization polynomial time, and concludes that many complex sets have polynomial-time implicit membership tests.

A Markov Decision Problem Approach to Goal Attainment A new Markov decision problem (MDP)-based method for managing goal attainment (GA), which is the process of planning and controlling actions that are related to the achievement of a set of defined goals in the presence of resource and time constraints, is proposed. Specifically, we address the problem as one of optimally selecting a sequence of actions to transform the system and/or its environment from an initial state to a desired state. We begin with a method of explicitly mapping an action-GA graph to an MDP graph and developing a dynamic programming (DP) recursion to solve the MDP problem. For larger problems having exponential complexity with respect to the number of goals, we propose guided search algorithms such as AO*, AOepsiv*, and greedy search techniques, whose search power rests on the efficiency of their heuristic evaluation functions (HEFs). Our contribution in this part stems from the introduction of a new problem-specific HEF to aid the search process. We demonstrate reductions in the computational costs of the proposed techniques through performance comparison with standard DP techniques. We conclude this paper with a method to address situations in which alternative strategies (e.g., second best) are required. The new extended AO* algorithm identifies alternative control sequences for attaining the organizational goals.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

FAWNdamentally power-efficient clusters

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

Parameterized complexity for the database theorist

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Kernel functions for case-based planning Case-based planning can take advantage of former problem-solving experiences by storing in a plan library previously generated plans that can be reused to solve similar planning problems in the future. Although comparative worst-case complexity analyses of plan generation and reuse techniques reveal that it is not possible to achieve provable efficiency gain of reuse over generation, we show that the case-based planning approach can be an effective alternative to plan generation when similar reuse candidates can be chosen. In this paper we describe an innovative case-based planning system, called OAKplan, which can efficiently retrieve planning cases from plan libraries containing more than ten thousand cases, choose heuristically a suitable candidate and adapt it to provide a good quality solution plan which is similar to the one retrieved from the case library. Given a planning problem we encode it as a compact graph structure, that we call Planning Encoding Graph, which gives us a detailed description of the topology of the planning problem. By using this graph representation, we examine an approximate retrieval procedure based on kernel functions that effectively match planning instances, achieving extremely good performance in standard benchmark domains. The experimental results point out the effect of the case base size and the importance of accurate matching functions for global system performance. Overall, we show that OAKplan is competitive with state-of-the-art plan generation systems in terms of number of problems solved, CPU time, plan difference values and plan quality when cases similar to the current planning problem are available in the plan library.

A review of machine learning for automated planning. Recent discoveries in automated planning are broadening the scope of planners, from toy problems to real applications. However, applying automated planners to real-world problems is far from simple. On the one hand, the definition of accurate action models for planning is still a bottleneck. On the other hand, off-the-shelf planners fail to scale-up and to provide good solutions in many domains. In these problematic domains, planners can exploit domain-specific control knowledge to improve their performance in terms of both speed and quality of the solutions. However, manual definition of control knowledge is quite difficult. This paper reviews recent techniques in machine learning for the automatic definition of planning knowledge. It has been organized according to the target of the learning process: automatic definition of planning action models and automatic definition of planning control knowledge. In addition, the paper reviews the advances in the related field of reinforcement learning.

Combining the Delete Relaxation with Critical-Path Heuristics: A Direct Characterization. Recent work has shown how to improve delete relaxation heuristics by computing relaxed plans, i. e., the hFF heuristic, in a compiled planning task ΠC which represents a given set C of fact conjunctions explicitly. While this compilation view of such partial delete relaxation is simple and elegant, its meaning with respect to the original planning task is opaque, and the size of ΠC grows exponentially in |C|. We herein provide a direct characterization, without compilation, making explicit how the approach arises from a combination of the delete-relaxation with critical-path heuristics. Designing equations characterizing a novel view on h+ on the one hand, and a generalized version hC of hm on the other hand, we show that h+(ΠC) can be characterized in terms of a combined hC+ equation. This naturally generalizes the standard delete-relaxation framework: understanding that framework as a relaxation over singleton facts as atomic subgoals, one can refine the relaxation by using the conjunctions C as atomic subgoals instead. Thanks to this explicit view, we identify the precise source of complexity in hFF(ΠC), namely maximization of sets of supported atomic subgoals during relaxed plan extraction, which is easy for singleton-fact subgoals but is NP-complete in the general case. Approximating that problem greedily, we obtain a polynomial-time hCFF version of hFF(ΠC), superseding the ΠC compilation, and superseding the modified ΠceC compilation which achieves the same complexity reduction but at an information loss. Experiments on IPC benchmarks show that these theoretical advantages can translate into empirical ones.

On the Complexity of Plan Adaptation by Derivational Analogy in a Universal Classical Planning Framework In this paper we present an algorithm called DerUCP, which can be regarded as a general model for plan adaptation using Derivational Analogy. Using DerUCP, we show that previous results on the complexity of plan adaptation do not apply to Derivational Analogy. We also show that Derivational Analogy can potentially produce exponential reductions in the size of the search space generated by a planning system.

Plan reuse versus plan generation: a theoretical and empirical analysis The ability of a planner to reuse parts of old plans is hypothesized to be a valuabletool for improving efficiency of planning by avoiding the repetition of the sameplanning effort. We test this hypothesis from an analytical and empirical point ofview. A comparative worst-case complexity analysis of generation and reuse underdifferent assumptions reveals that it is not possible to achieve a provable efficiencygain of reuse over generation. Further, assuming "conservative" plan...

An average case analysis of planning I present an average case analysis of propositional STRIPS planning. The analysis assumes that each possible precondition (likewise postcondition) is equally likely too appear within an operator. Under this assumption, I derive bounds for when it is highly likely that a planning instanee can be efficiently solved, either by finding a plan or proving that no plan exists. Roughly, if planning instances have no conditions (ground atoms), g goals, and O(n9√δ) operators, then a simple, efficient algorithm can prove that no plan exists for at least 1 - 8 of the instances. If instances have Ω(n(ln g)(ln g/δ)) operators, then a simple, efficient algorithm can find a plan for at least 1-δ of the instances. A similar result holds for plan modification, i.e., solving a planning instance that is close too another planning instance with a known plan. Thus it would appear that propositional STRIPS planning, a PSPACE-complete problem, is hard only for narrow parameter ranges, which complements previous average-case analyses for NP-complete problems. Future work is needed to narrow the gap between the bounds and to Consider more realistic distributional assumptions and more sophisticated algorithms.

Causal graphs and structurally restricted planning The causal graph is a directed graph that describes the variable dependencies present in a planning instance. A number of papers have studied the causal graph in both practical and theoretical settings. In this work, we systematically study the complexity of planning restricted by the causal graph. In particular, any set of causal graphs gives rise to a subcase of the planning problem. We give a complete classification theorem on causal graphs, showing that a set of graphs is either polynomial-time tractable, or is not polynomial-time tractable unless an established complexity-theoretic assumption fails; our theorem describes which graph sets correspond to each of the two cases. We also give a classification theorem for the case of reversible planning, and discuss the general direction of structurally restricted planning.

Hard and easy distributions of SAT problems We report results from large-scale experiments in satisfiability testing. As has been observed by others, testing the satisfiability of random formulas often appears surprisingly easy. Here we show that by using the right distribution of instances, and appropriate parameter values, it is possible to generate random formulas that are hard, that is, for which satisfiability testing is quite difficult. Our results provide a benchmark for the evaluation of satisfiability-testing procedures.

An Introduction to Least Commitment Planning Recent developments have clarified the process of generating partially ordered, partially specified sequences of actions whose execution will achieve an agent's goal. This article summarizes a progression of least commitment planners, starting with one that handles the simple STRIPS representation and ending with UCOPOP a planner that manages actions with disjunctive precondition, conditional effects, and universal quantification over dynamic universes. Along the way, I explain how Chapman's formulation of the modal truth criterion is misleading and why his NP-completeness result for reasoning about plans with conditional effects does not apply to UCOPOP.

Proceedings of the Sixteenth International Joint Conference on Artificial Intelligence, IJCAI 99, Stockholm, Sweden, July 31 - August 6, 1999. 2 Volumes, 1450 pages

The Astral Compendium For Protein Structure And Sequence Analysis The ASTRAL compendium provides several databases and tools to aid in the analysis of protein structures, particularly through the use of their sequences. The SPACI scores included in the system summarize the overall characteristics of a protein structure. A structural alignments database indicates residue equivalencies in superimposed protein domain structures, The PDB sequence-map files provide a linkage between the amino acid sequence of the molecule studied (SEQRES records in a database entry) and the sequence of the atoms experimentally observed in the structure (ATOM records). These maps are combined with information in the SCOP database to provide sequences of protein domains. Selected subsets of the domain database, with varying degrees of similarity measured in several different ways, are also available. ASTRAL may be accessed at http://astral.stanford.edu/.

Bidding for Storage Space in a Peer-to-Peer Data Preservation System Digital archives protect important data collections from failures by making multiple copies at other archives, so that there are always several good copies of a collection. In a cooperative replication network, sites "trade" space, so that each site contributes storage resources to the system and uses storage resources at other sites. Here, we examine bid trading: a mechanism where sites conduct auctions to determine who to trade with. A local site wishing to make a copy of a collection announces how much remote space is needed, and accepts bids for how much of its own space the local site must "pay" to acquire that remote space. We examine the best policies for determining when to call auctions and how much to bid, as well as the effects of "maverick" sites that attempt to subvert the bidding system. Simulations of auction and trading sessions indicate that bid trading can allowsites to achieve higher reliability than the alternative: a system where sites trade equal amounts of space without bidding.

A Conformant Planner with Explicit Disjunctive Representation of Belief States This paper describes a novel and competitive complete con- formant planner. Key to the enhanced performance is an effi- cient encoding of belief states as disjunctive normal form for- mulae and an efficient procedure for computing the successor belief state. We provide experimental comparative evaluation on a large pool of benchmarks. The novel design provides great efficiency and enhanced scalability, along with the intu- itive structure of disjunctive normal form representations.

Learning Topic Representation For Smt With Neural Networks Statistical Machine Translation (SMT) usually utilizes contextual information to disambiguate translation candidates. However, it is often limited to contexts within sentence boundaries, hence broader topical information cannot be leveraged. In this paper, we propose a novel approach to learning topic representation for parallel data using a neural network architecture, where abundant topical contexts are embedded via topic relevant monolingual data. By associating each translation rule with the topic representation, topic relevant rules are selected according to the distributional similarity with the source text during SMT decoding. Experimental results show that our method significantly improves translation accuracy in the NIST Chinese-to-English translation task compared to a state-of-the-art baseline.

Nonlinear autoassociation is not equivalent to PCA. A common misperception within the neural network community is that even with nonlinearities in their hidden layer, autoassociators trained with backpropagation are equivalent to linear methods such as principal component analysis (PCA). Our purpose is to demonstrate that nonlinear autoassociators actually behave differently from linear methods and that they can outperform these methods when used for latent extraction, projection, and classification. While linear autoassociators emulate PCA, and thus exhibit a flat or unimodal reconstruction error surface, autoassociators with nonlinearities in their hidden layer learn domains by building error reconstruction surfaces that, depending on the task, contain multiple local valleys. This interpolation bias allows nonlinear autoassociators to represent appropriate classifications of nonlinear multimodal domains, in contrast to linear autoassociators, which are inappropriate for such tasks. In fact, autoassociators with hidden unit nonlinearities can be shown to perform nonlinear classification and nonlinear recognition.

Transformation Invariant Autoassociation with Application to Handwritten Character Recognition When training neural networks by the classical backpropagation algo- rithm the whole problem to learn must be expressed by a set of inputs and desired outputs. However, we often have high-level knowledge about the learning problem. In optical character recognition (OCR), for in- stance, we know that the classification should be invariant under a set of transformations like rotation or translation. We propose a new mo dular classification system based on several autoassociative multilayer percep- trons which allows the efficient incorporation of such knowledge. Results are reported on the NIST database of upper case handwritten letters and compared to other approaches to the invariance problem.

Rational Kernels: Theory and Algorithms Many classification algorithms were originally designed for fixed-size vectors. Recent applications in text and speech processing and computational biology require however the analysis of variable-length sequences and more generally weighted automata. An approach widely used in statistical learning techniques such as Support Vector Machines (SVMs) is that of kernel methods, due to their computational efficiency in high-dimensional feature spaces. We introduce a general family of kernels based on weighted transducers or rational relations, rational kernels , that extend kernel methods to the analysis of variable-length sequences or more generally weighted automata. We show that rational kernels can be computed efficiently using a general algorithm of composition of weighted transducers and a general single-source shortest-distance algorithm. Not all rational kernels are positive definite and symmetric (PDS), or equivalently verify the Mercer condition, a condition that guarantees the convergence of training for discriminant classification algorithms such as SVMs. We present several theoretical results related to PDS rational kernels. We show that under some general conditions these kernels are closed under sum, product, or Kleene-closure and give a general method for constructing a PDS rational kernel from an arbitrary transducer defined on some non-idempotent semirings. We give the proof of several characterization results that can be used to guide the design of PDS rational kernels. We also show that some commonly used string kernels or similarity measures such as the edit-distance, the convolution kernels of Haussler, and some string kernels used in the context of computational biology are specific instances of rational kernels. Our results include the proof that the edit-distance over a non-trivial alphabet is not negative definite, which, to the best of our knowledge, was never stated or proved before. Rational kernels can be combined with SVMs to form efficient and powerful techniques for a variety of classification tasks in text and speech processing, or computational biology. We describe examples of general families of PDS rational kernels that are useful in many of these applications and report the result of our experiments illustrating the use of rational kernels in several difficult large-vocabulary spoken-dialog classification tasks based on deployed spoken-dialog systems. Our results show that rational kernels are easy to design and implement and lead to substantial improvements of the classification accuracy.

Describing Visual Scenes Using Transformed Objects and Parts We develop hierarchical, probabilistic models for objects, the parts composing them, and the visual scenes surrounding them. Our approach couples topic models originally developed for text analysis with spatial transformations, and thus consistently accounts for geometric constraints. By building integrated scene models, we may discover contextual relationships, and better exploit partially labeled training images. We first consider images of isolated objects, and show that sharing parts among object categories improves detection accuracy when learning from few examples. Turning to multiple object scenes, we propose nonparametric models which use Dirichlet processes to automatically learn the number of parts underlying each object category, and objects composing each scene. The resulting transformed Dirichlet process (TDP) leads to Monte Carlo algorithms which simultaneously segment and recognize objects in street and office scenes.

Statistical models for partial membership We present a principled Bayesian framework for modeling partial memberships of data points to clusters. Unlike a standard mixture model which assumes that each data point belongs to one and only one mixture component, or cluster, a partial membership model allows data points to have fractional membership in multiple clusters. Algorithms which assign data points partial memberships to clusters can be useful for tasks such as clustering genes based on microarray data (Gasch & Eisen, 2002). Our Bayesian Partial Membership Model (BPM) uses exponential family distributions to model each cluster, and a product of these distibtutions, with weighted parameters, to model each datapoint. Here the weights correspond to the degree to which the datapoint belongs to each cluster. All parameters in the BPM are continuous, so we can use Hybrid Monte Carlo to perform inference and learning. We discuss relationships between the BPM and Latent Dirichlet Allocation, Mixed Membership models, Exponential Family PCA, and fuzzy clustering. Lastly, we show some experimental results and discuss nonparametric extensions to our model.

The Entire Regularization Path for the Support Vector Machine The support vector machine (SVM) is a widely used tool for classification. Many efficient implementations exist for fitting a two-class SVM model. The user has to supply values for the tuning parameters: the regularization cost parameter, and the kernel parameters. It seems a common practice is to use a default value for the cost parameter, often leading to the least restrictive model. In this paper we argue that the choice of the cost parameter can be critical. We then derive an algorithm that can fit the entire path of SVM solutions for every value of the cost parameter, with essentially the same computational cost as fitting one SVM model. We illustrate our algorithm on some examples, and use our representation to give further insight into the range of SVM solutions.

Learning nonlinear overcomplete representations for efficient coding We derive a learning algorithm for inferring an overcomplete basisby viewing it as probabilistic model of the observed data. Overcompletebases allow for better approximation of the underlyingstatistical density. Using a Laplacian prior on the basis coefficientsremoves redundancy and leads to representations that are sparseand are a nonlinear function of the data. This can be viewed asa generalization of the technique of independent component analysisand provides a method for blind ...

On Kernel-Target Alignment We introduce the notion of kernel-alignment, a measure of similarity between two kernel functions or between a kernel and a target function. This quantity captures the degree of agreement between a kernel and a given learning task, and has very natural interpretations in machine learning, leading also to simple algorithms for model selection and learning. We analyse its theoretical properties, proving that it is sharply concentrated around its expected value, and we discuss its relation with other standard measures of performance. Finally we describe some of the algorithms that can be obtained within this framework, giving experimental results showing that adapting the kernel to improve alignment on the labelled data significantly increases the alignment on the test set, giving improved classification accuracy. Hence, the approach provides a principled method of performing transduction.

Support-Vector Networks The support-vector network is a new learning machine for two-group classification problems. The machine conceptually implements the following idea: input vectors are non-linearly mapped to a very high-dimension feature space. In this feature space a linear decision surface is constructed. Special properties of the decision surface ensures high generalization ability of the learning machine. The idea behind the support-vector network was previously implemented for the restricted case where the training data can be separated without errors. We here extend this result to non-separable training data.High generalization ability of support-vector networks utilizing polynomial input transformations is demonstrated. We also compare the performance of the support-vector network to various classical learning algorithms that all took part in a benchmark study of Optical Character Recognition.

An efficient learning procedure for deep Boltzmann machines. We present a new learning algorithm for Boltzmann machines that contain many layers of hidden variables. Data-dependent statistics are estimated using a variational approximation that tends to focus on a single mode, and data-independent statistics are estimated using persistent Markov chains. The use of two quite different techniques for estimating the two types of statistic that enter into the gradient of the log likelihood makes it practical to learn Boltzmann machines with multiple hidden layers and millions of parameters. The learning can be made more efficient by using a layer-by-layer pretraining phase that initializes the weights sensibly. The pretraining also allows the variational inference to be initialized sensibly with a single bottom-up pass. We present results on the MNIST and NORB data sets showing that deep Boltzmann machines learn very good generative models of handwritten digits and 3D objects. We also show that the features discovered by deep Boltzmann machines are a very effective way to initialize the hidden layers of feedforward neural nets, which are then discriminatively fine-tuned.

Sentiment classification based on supervised latent n-gram analysis In this paper, we propose an efficient embedding for modeling higher-order (n-gram) phrases that projects the n-grams to low-dimensional latent semantic space, where a classification function can be defined. We utilize a deep neural network to build a unified discriminative framework that allows for estimating the parameters of the latent space as well as the classification function with a bias for the target classification task at hand. We apply the framework to large-scale sentimental classification task. We present comparative evaluation of the proposed method on two (large) benchmark data sets for online product reviews. The proposed method achieves superior performance in comparison to the state of the art.

Logic Programming and Negation: A Survey. We survey here various approaches which were proposed to incorporate negation in logicprograms. We concentrate on the proof-theoretic and model-theoretic issues and the relationshipsbetween them.1991 Mathematics Subject Classification: 68Q40, 68T15.CR Categories: F.3.2., F.4.1, H.3.3, I.2.3.Keywords and Phrases: negation, general logic programs, non-monotonic reasoning.Notes. The work of the first author was partly supported by ESPRIT Basic Research Action6810 (Compulog 2). The work...

Planning with sensing, concurrency, and exogenous events: logical framework and implementation The focus of current research in cognitive robotics is both on the realization of sys- tems based on known formal settings and on the extension of previous formal approaches to account for features that play a signifl- cant role for autonomous robots, but have not yet received an adequate treatment. In this paper we adopt a formal framework de- rived from Propositional Dynamic Logics by exploiting their formal correspondence with Description Logics, and present an extension of such a framework obtained by introducing both concurrency on primitive actions and autoepistemic operators for explicitly repre- senting the robot's epistemic state. We show that the resulting formal setting allows for the representation of actions with context- dependent efiects, sensing actions, and con- current actions, and address both the pres- ence of exogenous events and the characteri- zation of the notion of executable plan in such a complex setting. Moreover, we present an implementation of this framework in a system which is capable of generating plans that are actually executed on mobile robots, and illus- trate the experimentation of such a system in the design and implementation of soccer players for the 1999 Robocup competition.

Editorial introduction to the Neural Networks special issue on Deep Learning of Representations.

Evolving neural networks for strategic decision-making problems. Evolution of neural networks, or neuroevolution, has been a successful approach to many low-level control problems such as pole balancing, vehicle control, and collision warning. However, certain types of problems–such as those involving strategic decision-making–have remained difficult for neuroevolution to solve. This paper evaluates the hypothesis that such problems are difficult because they are fractured: The correct action varies discontinuously as the agent moves from state to state. A method for measuring fracture using the concept of function variation is proposed and, based on this concept, two methods for dealing with fracture are examined: neurons with local receptive fields, and refinement based on a cascaded network architecture. Experiments in several benchmark domains are performed to evaluate how different levels of fracture affect the performance of neuroevolution methods, demonstrating that these two modifications improve performance significantly. These results form a promising starting point for expanding neuroevolution to strategic tasks.

A training algorithm for optimal margin classifiers A training algorithm that maximizes the margin between the training patterns and the decision boundary is presented. The technique is applicable to a wide variety of the classification functions, including Perceptrons, polynomials, and Radial Basis Functions. The effective number of parameters is adjusted automatically to match the complexity of the problem. The solution is expressed as a linear combination of supporting patterns. These are the subset of training patterns that are closest to the decision boundary. Bounds on the generalization performance based on the leave-one-out method and the VC-dimension are given. Experimental results on optical character recognition problems demonstrate the good generalization obtained when compared with other learning algorithms.

Learning Deep Architectures for AI Theoretical results suggest that in order to learn the kind of com-plicated functions that can represent high-level abstractions (e.g., invision, language, and other AI-level tasks), one may needdeep architec-tures. Deep architectures are composed of multiple levels of non-linearoperations, such as in neural nets with many hidden layers or in com-plicated propositional formulae re-using many sub-formulae. Searchingthe parameter space of deep architectures is a difficult task, but learningalgorithms such as those for Deep Belief Networks have recently beenproposed to tackle this problem with notable success, beating the state-of-the-art in certain areas. This monograph discusses the motivationsand principles regarding learning algorithms for deep architectures, inparticular those exploiting as building blocks unsupervised learning ofsingle-layer models such as Restricted Boltzmann Machines, used toconstruct deeper models such as Deep Belief Networks.

Extended stable semantics for normal and disjunctive programs

The nature of statistical learning theory~. First Page of the Article

A machine program for theorem-proving The programming of a proof procedure is discussed in connection with trial runs and possible improvements.

An Introduction to Least Commitment Planning Recent developments have clarified the process of generating partially ordered, partially specified sequences of actions whose execution will achieve an agent's goal. This article summarizes a progression of least commitment planners, starting with one that handles the simple STRIPS representation and ending with UCOPOP a planner that manages actions with disjunctive precondition, conditional effects, and universal quantification over dynamic universes. Along the way, I explain how Chapman's formulation of the modal truth criterion is misleading and why his NP-completeness result for reasoning about plans with conditional effects does not apply to UCOPOP.

Equilibria and steering laws for planar formations This paper presents a Lie group setting for the problem of control of formations, as a natural outcome of the analysis of a planar two-vehicle formation control law. The vehicle trajectories are described using the planar Frenet–Serret equations of motion, which capture the evolution of both the vehicle position and orientation for unit-speed motion subject to curvature (steering) control. The set of all possible (relative) equilibria for arbitrary G-invariant curvature controls is described (where G=SE(2) is a symmetry group for the control law), and a global convergence result for the two-vehicle control law is proved. An n-vehicle generalization of the two-vehicle control law is also presented, and the corresponding (relative) equilibria for the n-vehicle problem are characterized. Work is on-going to discover stability and convergence results for the n-vehicle problem.

Improving the I/O Performance of Real-Time Database Systems with Multiple-Disk Storage Structures

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Egocentric Video Biometrics.

Stacked Denoising Autoencoders: Learning Useful Representations in a Deep Network with a Local Denoising Criterion We explore an original strategy for building deep networks, based on stacking layers of denoising autoencoders which are trained locally to denoise corrupted versions of their inputs. The resulting algorithm is a straightforward variation on the stacking of ordinary autoencoders. It is however shown on a benchmark of classification problems to yield significantly lower classification error, thus bridging the performance gap with deep belief networks (DBN), and in several cases surpassing it. Higher level representations learnt in this purely unsupervised fashion also help boost the performance of subsequent SVM classifiers. Qualitative experiments show that, contrary to ordinary autoencoders, denoising autoencoders are able to learn Gabor-like edge detectors from natural image patches and larger stroke detectors from digit images. This work clearly establishes the value of using a denoising criterion as a tractable unsupervised objective to guide the learning of useful higher level representations.

Learning Deep Architectures for AI Theoretical results suggest that in order to learn the kind of com-plicated functions that can represent high-level abstractions (e.g., invision, language, and other AI-level tasks), one may needdeep architec-tures. Deep architectures are composed of multiple levels of non-linearoperations, such as in neural nets with many hidden layers or in com-plicated propositional formulae re-using many sub-formulae. Searchingthe parameter space of deep architectures is a difficult task, but learningalgorithms such as those for Deep Belief Networks have recently beenproposed to tackle this problem with notable success, beating the state-of-the-art in certain areas. This monograph discusses the motivationsand principles regarding learning algorithms for deep architectures, inparticular those exploiting as building blocks unsupervised learning ofsingle-layer models such as Restricted Boltzmann Machines, used toconstruct deeper models such as Deep Belief Networks.

Extended stable semantics for normal and disjunctive programs

The nature of statistical learning theory~. First Page of the Article

A machine program for theorem-proving The programming of a proof procedure is discussed in connection with trial runs and possible improvements.

An Introduction to Least Commitment Planning Recent developments have clarified the process of generating partially ordered, partially specified sequences of actions whose execution will achieve an agent's goal. This article summarizes a progression of least commitment planners, starting with one that handles the simple STRIPS representation and ending with UCOPOP a planner that manages actions with disjunctive precondition, conditional effects, and universal quantification over dynamic universes. Along the way, I explain how Chapman's formulation of the modal truth criterion is misleading and why his NP-completeness result for reasoning about plans with conditional effects does not apply to UCOPOP.

Equilibria and steering laws for planar formations This paper presents a Lie group setting for the problem of control of formations, as a natural outcome of the analysis of a planar two-vehicle formation control law. The vehicle trajectories are described using the planar Frenet–Serret equations of motion, which capture the evolution of both the vehicle position and orientation for unit-speed motion subject to curvature (steering) control. The set of all possible (relative) equilibria for arbitrary G-invariant curvature controls is described (where G=SE(2) is a symmetry group for the control law), and a global convergence result for the two-vehicle control law is proved. An n-vehicle generalization of the two-vehicle control law is also presented, and the corresponding (relative) equilibria for the n-vehicle problem are characterized. Work is on-going to discover stability and convergence results for the n-vehicle problem.

Pushing the envelope: planning, propositional logic, and stochastic search Planning is a notoriously hard combinatorial search problem. In many interesting domains, current planning algorithms fail to scale up gracefully. By combining a general, stochastic search algorithm and appropriate problem encodings based on propositional logic, we are able to solve hard planning problems many times faster than the best current planning systems. Although stochastic methods have been shown to be very effective on a wide range of scheduling problems, this is the first demonstration of its power on truly challenging classical planning instances. This work also provides a new perspective on representational issues in planning.

Simultaneous Localization And Mapping With Sparse Extended Information Filters In this paper we describe a scalable algorithm for the simultaneous mapping and localization (SLAM) problem. SLAM is the problem of acquiring a map of a static environment with a mobile robot. The vast majority of SLAM algorithms are based on the extended Kahnan filter (EKF), In this paper we advocate an algorithm that relies on the dual of the EKE the extended information filter (EIF). We show that when represented in the information form, map posteriors are dominated by a small number of links that tie together nearby features in the map. This insight is developed into a sparse variant of the EIF called the sparse extended information filter (SEIF). SEIFs represent maps by graphical networks of features that are locally interconnected, where links represent relative information between pairs of nearby,features, as well as information about the robot's pose relative to the map. We show that all essential update equations in SEIFs can be executed in constant time, irrespective of the size of the map. We also provide empirical results obtained for a benchmark data set collected in an outdoor environment, and using a multi-robot mapping simulation.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Learning deep structured semantic models for web search using clickthrough data Latent semantic models, such as LSA, intend to map a query to its relevant documents at the semantic level where keyword-based matching often fails. In this study we strive to develop a series of new latent semantic models with a deep structure that project queries and documents into a common low-dimensional space where the relevance of a document given a query is readily computed as the distance between them. The proposed deep structured semantic models are discriminatively trained by maximizing the conditional likelihood of the clicked documents given a query using the clickthrough data. To make our models applicable to large-scale Web search applications, we also use a technique called word hashing, which is shown to effectively scale up our semantic models to handle large vocabularies which are common in such tasks. The new models are evaluated on a Web document ranking task using a real-world data set. Results show that our best model significantly outperforms other latent semantic models, which were considered state-of-the-art in the performance prior to the work presented in this paper.

Learning Continuous Phrase Representations For Translation Modeling This paper tackles the sparsity problem in estimating phrase translation probabilities by learning continuous phrase representations, whose distributed nature enables the sharing of related phrases in their representations. A pair of source and target phrases are projected into continuous-valued vector representations in a low-dimensional latent space, where their translation score is computed by the distance between the pair in this new space. The projection is performed by a neural network whose weights are learned on parallel training data. Experimental evaluation has been performed on two WMT translation tasks. Our best result improves the performance of a state-of-the-art phrase-based statistical machine translation system trained on WMT 2012 French-English data by up to 1.3 BLEU points.

Audio Chord Recognition with Recurrent Neural Networks.

Using Recurrent Neural Networks for Slot Filling in Spoken Language Understanding Semantic slot filling is one of the most challenging problems in spoken language understanding (SLU). In this paper, we propose to use recurrent neural networks (RNNs) for this task, and present several novel architectures designed to efficiently model past and future temporal dependencies. Specifically, we implemented and compared several important RNN architectures, including Elman, Jordan, and hybrid variants. To facilitate reproducibility, we implemented these networks with the publicly available Theano neural network toolkit and completed experiments on the well-known airline travel information system (ATIS) benchmark. In addition, we compared the approaches on two custom SLU data sets from the entertainment and movies domains. Our results show that the RNN-based models outperform the conditional random field (CRF) baseline by 2% in absolute error reduction on the ATIS benchmark. We improve the state-of-the-art by 0.5% in the Entertainment domain, and 6.7% for the movies domain.

Parametric Learning of Deep Convolutional Neural Network Deep neural networks have recently been showing great potential on visual recognition tasks. However, it is also considered difficult to tune its parameters, and it has high training cost. This work focuses on analysis of several learning methods and properties of multinomial logistic regression deep convolutional network. We implemented a scalable deep neural network, compared the efficiency of different methods and how parameters affect the learning process. We propose an efficient method of performing back-propagation with limited kernel functions on GPU and achieved better efficiency. Our conclusions can be applied to train deep networks more efficiently. We achieved the recognition rate of over 0.95 without image preprocessing and fine tuning, within 10 minutes on a single machine.

Preliminary investigation of Boltzmann machine classifiers for speaker recognition.

Learning Deep Energy Models.

Automatic Identification of Instrument Classes in Polyphonic and Poly-Instrument Audio.

Semantic text classification of disease reporting Traditional text classification studied in the IR literature is mainly based on topics. That is, each class or category represents a particular topic, e.g., sports, politics or sciences. However, many real-world text classification problems require more refined classification based on some semantic aspects. For example, in a set of documents about a particular disease, some documents may report the outbreak of the disease, some may describe how to cure the disease, some may discuss how to prevent the disease, and yet some others may include all the above information. To classify text at this semantic level, the traditional "bag of words" model is no longer sufficient. In this paper, we report a text classification study at the semantic level and show that sentence semantic and structure features are very useful for such kind of classification. Our experimental results based on a disease outbreak dataset demonstrated the effectiveness of the proposed approach.

Training Deep Convolutional Neural Networks to Play Go. Mastering the game of Go has remained a long standing challenge to the field of AI. Modern computer Go systems rely on processing millions of possible future positions to play well, but intuitively a stronger and more 'humanlike' way to play the game would be to rely on pattern recognition abilities rather then brute force computation. Following this sentiment, we train deep convolutional neural networks to play Go by training them to predict the moves made by expert Go players. To solve this problem we introduce a number of novel techniques, including a method of tying weights in the network to 'hard code' symmetries that are expect to exist in the target function, and demonstrate in an ablation study they considerably improve performance. Our final networks are able to achieve move prediction accuracies of 41.1% and 44.4% on two different Go datasets, surpassing previous state of the art on this task by significant margins. Additionally, while previous move prediction programs have not yielded strong Go playing programs, we show that the networks trained in this work acquired high levels of skill. Our convolutional neural networks can consistently defeat the well known Go program GNU Go, indicating it is state of the art among programs that do not use Monte Carlo Tree Search. It is also able to win some games against state of the art Go playing program Fuego while using a fraction of the play time. This success at playing Go indicates high level principles of the game were learned.

Hot Mirroring: A Study to Hide Parity Upgrade Penalty and Degradations During Rebuilds for RAID5

Data exchange: getting to the core Data exchange is the problem of taking data structured under a source schema and creating an instance of a target schema that reflects the source data as accurately as possible. Given a source instance, there may be many solutions to the data exchange problem, that is, many target instances that satisfy the constraints of the data exchange problem. In an earlier article, we identified a special class of solutions that we call universal. A universal solution has homomorphisms into every possible solution, and hence is a “most general possible” solution. Nonetheless, given a source instance, there may be many universal solutions. This naturally raises the question of whether there is a “best” universal solution, and hence a best solution for data exchange. We answer this question by considering the well-known notion of the core of a structure, a notion that was first studied in graph theory, and has also played a role in conjunctive-query processing. The core of a structure is the smallest substructure that is also a homomorphic image of the structure. All universal solutions have the same core (up to isomorphism); we show that this core is also a universal solution, and hence the smallest universal solution. The uniqueness of the core of a universal solution together with its minimality make the core an ideal solution for data exchange. We investigate the computational complexity of producing the core. Well-known results by Chandra and Merlin imply that, unless P = NP, there is no polynomial-time algorithm that, given a structure as input, returns the core of that structure as output. In contrast, in the context of data exchange, we identify natural and fairly broad conditions under which there are polynomial-time algorithms for computing the core of a universal solution. We also analyze the computational complexity of the following decision problem that underlies the computation of cores: given two graphs G and H, is H the core of G? Earlier results imply that this problem is both NP-hard and coNP-hard. Here, we pinpoint its exact complexity by establishing that it is a DP-complete problem. Finally, we show that the core is the best among all universal solutions for answering existential queries, and we propose an alternative semantics for answering queries in data exchange settings.

Domain-dependent knowledge in answer set planning In this article we consider three different kinds of domain-dependent control knowledge (temporal, procedural and HTN-based) that are useful in planning. Our approach is declarative and relies on the language of logic programming with answer set semantics (AnsProlog*). AnsProlog* is designed to plan without control knowledge. We show how temporal, procedural and HTN-based control knowledge can be incorporated into AnsProlog* by the modular addition of a small number of domain-dependent rules, without the need to modify the planner. We formally prove the correctness of our planner, both in the absence and presence of the control knowledge. Finally, we perform some initial experimentation that demonstrates the potential reduction in planning time that can be achieved when procedural domain knowledge is used to solve planning problems with large plan length.

On Qualitative Route Descriptions. The generation of route descriptions is a fundamental task of navigation systems. A particular problem in this context is to identify routes that can easily be described and processed by users. In this work, we present a framework for representing route networks with the qualitative information necessary to evaluate and optimize route descriptions with regard to ambiguities in them. We identify different agent models that differ in how agents are assumed to process route descriptions while navigating through route networks and discuss which agent models can be translated into PDL programs. Further, we analyze the computational complexity of matching route descriptions and paths in route networks in dependency of the agent model. Finally, we empirically evaluate the influence of the agent model on the optimization and the processing of route instructions.

Combination Of Two-Dimensional Cochleogram And Spectrogram Features For Deep Learning-Based Asr This paper explores the use of auditory features based on cochleograms; two dimensional speech features derived from gammatone filters within the convolutional neural network (CNN) framework. Furthermore, we also propose various possibilities to combine cochleogram features with log-mel filter banks or spectrogram features. In particular, we combine within low and high levels of CNN framework which we refer to as low-level and high-level feature combination. As comparison, we also construct the similar configuration with deep neural network (DNN). Performance was evaluated in the framework of hybrid neural network - hidden Markov model (NN-HMM) system on TIMIT phoneme sequence recognition task. The results reveal that cochleogram-spectrogram feature combination provides significant advantages. The best accuracy was obtained by high-level combination of two dimensional cochleogram-spectrogram features using CNN, achieved up to 8.2% relative phoneme error rate (PER) reduction from CNN single features or 19.7% relative PER reduction from DNN single features.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Learning Topic Representation For Smt With Neural Networks Statistical Machine Translation (SMT) usually utilizes contextual information to disambiguate translation candidates. However, it is often limited to contexts within sentence boundaries, hence broader topical information cannot be leveraged. In this paper, we propose a novel approach to learning topic representation for parallel data using a neural network architecture, where abundant topical contexts are embedded via topic relevant monolingual data. By associating each translation rule with the topic representation, topic relevant rules are selected according to the distributional similarity with the source text during SMT decoding. Experimental results show that our method significantly improves translation accuracy in the NIST Chinese-to-English translation task compared to a state-of-the-art baseline.

Disjunctive signed logic programs In this work, we define signed disjunctive programs and investigate the existence of answer sets for this class of programs. Our main argument is based on an analogue of Tarski's fixed point theorem which we prove for multivalued mappings. This is an original approach compared to known techniques used to prove the existence of answer sets for disjunctive programs.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

Parameterized complexity for the database theorist

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

On the Complexity of Semantic Self-minimization Partial Kripke structures model only parts of a state space and so enable aggressive abstraction of systems prior to verifying them with respect to a formula of temporal logic. This partiality of models means that verifications may reply with true (all refinements satisfy the formula under check), false (no refinement satisfies the formula under check) or don't know. Generalized model checking is the most precise verification for such models (all don't know answers imply that some refinements satisfy the formula, some don't), but computationally expensive. A compositional model-checking algorithm for partial Kripke structures is efficient, sound (all answers true and false are truthful), but may lose precision by answering don't know instead of a factual true or false. Recent work has shown that such a loss of precision does not occur for this compositional algorithm for most practically relevant patterns of temporal logic formulas. Formulas that never lose precision in this manner are called semantically self-minimizing. In this paper we provide a systematic study of the complexity of deciding whether a formula of propositional logic, propositional modal logic or the propositional modal mu-calculus is semantically self-minimizing.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Exploring Gate-Limited Analytical Models for High Performance Network Storage Servers

Parameterized complexity for the database theorist

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

RSN1-tt(NP) Distinguishes Robust Many-One and Turing Completeness Do complexity classes have many-one complete sets if and only if they have Turing- complete sets? We prove that there is a relativized world in which a relatively natural complexity class—namely a downward closure of NP, RSN1-tt(NP)—has Turing-complete sets but has no many-one complete sets. In fact, we show that in the same relativized world this class has 2-truth-table complete sets but lacks 1-truth-table complete sets. As part of the groundwork for our result, we prove that RSN1-tt(NP) has many equivalent forms having to do with ordered and parallel access to NP and NP ∩ coNP.

Pinpointing Computation with Modular Queries in the Boolean Hierarchy A modular query consists of asking how many (modulo m) of k strings belong to a fixed NP language. Modular queries provide a form of restricted access to an NP oracle. For each k and m, we consider the class of languages accepted by NP machines that ask a single modular query. Han and Thierauf (HT95) showed that these classes coincide with levels of the Boolean hierarchy when m is even or k 2m, and they determined the exact levels. Until now, the remaining case — odd m and large k — looked quite dicult. We pinpoint the level in the Boolean hierarchy for the remaining case; thus, these classes coincide with levels of the Boolean hierarchy for every k and m. In addition we characterize the classes obtained by using an NP(l) acceptor in place of an NP acceptor (NP(l) is the lth level of the Boolean hierarchy). As before, these all coincide with levels in the Boolean hierarchy.

Query Order in the Polynomial Hierarchy We study query order within the polynomial hierarchy. $P^{\cal C : \cal D}$ denotes the class of languages computable by a polynomial-time machine that is allowed one query to $\cal C$ followed by one query to $\cal D$. We prove that the levels of the polynomial hierarchy are order-oblivious: $P^{\sum^p_j:\sum^p_k} = P^{\sum^p_k:\sum^p_j}. Yet, we also show that these ordered query classes form new levels in the polynomial hierarchy unless the polynomial hierarchy collapses. We prove that all leaf language classes---and thus essentially all standard complexity classes---inherit all order-obliviousness results that hold for P.

A Downward Collapse within the Polynomial Hierarchy Downward collapse (also known as upward separation) refers to cases where the equality of two larger classes implies the equality of two smaller classes. We provide an unqualified downward collapse result completely within the polynomial hierarchy. In particular, we prove that, for k 2, if ${\rm P}^{\Sigma^p_k[1]} = {\rm P}^{\Sigma^p_k[2]}$ then $\Sigma^p_k = \Pi^p_k = {\rm PH}$. We extend this to obtain a more general downward collapse result.

On boolean lowness and boolean highness The concepts of lowness and highness originate from recursion theory and were introduced into the complexity theory by Schoning (Lecture Notes in Computer Science, Vol. 211, Springer, Berlin, 1985). Informally, a set is low (high resp.) for a relativizable class K of languages if it does not add (adds maximal resp.) power to K when used as an oracle. In this paper, we introduce the notions of boolean lowness and boolean highness. Informally, a set is boolean low (boolean high resp.) for a class X of languages if it does not add (adds maximal resp.) power to K when combined with K by boolean operations. We prove properties of boolean lowness and boolean highness which show a lot of similarities with the notions of lowness and highness. Using Kadin's technique of hard strings (see Kadin, SIAM J. Comput 17(6) (1988) 1263-1282; Wagner, Number-of-query hierachies, TR 158, University of Augsburg, 1987; Chang and Kadin SIAM J. Comput. 25(2) (1996) 340; Beigel ct al. Math. Systems Theory 26 (1993) 293-310) we show that the sets which are boolean low for the classes of the boolean hierarchy are low for the boolean closure of Sigma (p)(2). Furthermore, we prove a result on boolean lowness which has as a corollary the best known result (sec Beigel, (1993); in fact even a bit better) on the connection of the collapses of the boolean hierarchy and the polynomial-lime hierarchy if BH = NP(L) then PH = Sigma (p)(2)(k - 1) circle plus NP(k). (C) 2001 Published by Elsevier Science B.V.

Bounded queries, approximations, and the Boolean hierarchy This paper investigates nondeterministic bounded query classes in relation to the complexity of NP-hard approximation problems and the Boolean Hierarchy. Nondeterministic bounded query classes turn out to be rather suitable for describing the complexity of NP-hard approximation problems. The results in this paper take advantage of this machine-based.

The complexity of promise problems with applications to public-key cryptography

Exact Complexity of Exact-Four-Colorability and of the Winner Problem for Young Elections We classify two problems: Exact-Four-Colorability and the winner problem for Young elections. Regarding the former problem, Wagner raised the question of whether it is DP-complete to determine if the chromatic number of a given graph is exactly four. We prove a general result that in particular solves Wagner's question in the affirmative. In 1977, Young proposed a voting scheme that extends the Condorcet Principle based on the fewest possible number of voters whose removal yields a Condorcet winner. We prove that both the winner and the ranking problem for Young elections is complete for p(parallel to)(NP), the class of problems solvable in polynomial time by parallel access to NP. Analogous results for Lewis Carroll's 1876 voting scheme were recently established by Hemaspaandra et al. In contrast, we prove that the winner and ranking problems in Fishburn's homogeneous variant of Carroll's voting scheme can be solved efficiently by linear programming.

A Goal-Oriented Approach to Computing Well Founded Semantics

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

The complexity of Markov decision processes We investigate the complexity of the classical problem of optimal policy computation in Markov decision processes. All three variants of the problem finite horizon, infinite horizon discounted, and...

Generalized working sets for segment reference strings The working-set concept is extended for programs that reference segments of different sizes. The generalized working-set policy (GWS) keeps as its resident set those segments whose retention costs do not exceed their retrieval costs. The GWS is a model for the entire class of demand-fetching memory policies that satisfy a resident-set inclusion property. A generalized optimal policy (GOPT) is also defined; at its operating points it minimizes aggregated retention and swapping costs. Special cases of the cost structure allow GWS and GOPT to simulate any known stack algorithm, the working set, and VMIN. Efficient procedures for computing demand curves showing swapping load as a function of memory usage are developed for GWS and GOPT policies. Empirical data from an actual system are included.

Global reinforcement learning in neural networks. In this letter, we have found a more general formulation of the REward Increment = Nonnegative Factor x Offset Reinforcement x Characteristic Eligibility (REINFORCE) learning principle first suggested by Williams. The new formulation has enabled us to apply the principle to global reinforcement learning in networks with various sources of randomness, and to suggest several simple local rules for such networks. Numerical simulations have shown that for simple classification and reinforcement learning tasks, at least one family of the new learning rules gives results comparable to those provided by the famous Rules A(r-i) and A(r-p) for the Boltzmann machines.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Treating Highly Anisotropic Subsurface Flow with the Multiscale Finite-Volume Method The multiscale finite-volume (MSFV) method has been designed to solve flow problems on large domains efficiently. First, a set of basis functions, which are local numerical solutions, is employed to construct a fine-scale pressure approximation; then a conservative fine-scale velocity approximation is constructed by solving local problems with boundary conditions obtained from the pressure approximation; finally, transport is solved at the. ne scale. The method proved very robust and accurate for multiphase flow simulations in highly heterogeneous isotropic reservoirs with complex correlation structures. However, it has recently been pointed out that the fine-scale details of the MSFV solutions may be lost in the case of high anisotropy or large grid aspect ratios. This shortcoming is analyzed in this paper, and it is demonstrated that it is caused by the appearance of unphysical "circulation cells." We show that damped-shear boundary conditions for the conservative-velocity problems or linear boundary conditions for the basis-function problems can significantly improve the MSFV solution for highly anisotropic permeability fields without sensitively affecting the solution in the isotropic case.

Analysis of two-scale finite volume element method for elliptic problem In this paper we propose and analyze a class of finite volume element method for solving a second order elliptic boundary value problem whose solution is defined in more than one length scales. The method has the ability to incorporate the small scale behaviors of the solution on the large scale one. This is achieved through the construction of the basis functions on each element that satisfy the homogeneous elliptic differential equation. Furthermore, the method enjoys numerical conservation feature which is highly desirable in many applications. Existing analyses on its finite element counterpart reveal that there exists a resonance error between the mesh size and the small length scale. This result motivates an oversampling technique to overcome this drawback. We develop an analysis of the proposed method under the assumption that the coefficients are of two scales and periodic in the small scale. The theoretical results are confirmed experimentally by several convergence tests. Moreover, we present an application of the method to flows in porous media.

Compact Multiscale Finite Volume Method for Heterogeneous Anisotropic Elliptic Equations The multiscale finite volume (MSFV) method is introduced for the efficient solution of elliptic problems with rough coefficients in the absence of scale separation. The coarse operator of the MSFV method is presented as a multipoint flux approximation (MPFA) with numerical evaluation of the transmissibilities. The monotonicity region of the original MSFV coarse operator has been determined for the homogeneous anisotropic case. For grid-aligned anisotropy the monotonicity of the coarse operator is very limited. A compact coarse operator for the MSFV method is presented that reduces to a 7-point stencil with optimal monotonicity properties in the homogeneous case. For heterogeneous cases the compact coarse operator improves the monotonicity of the MSFV method, especially for anisotropic problems. The compact operator also leads to a coarse linear system much closer to an M-matrix. Gradients in the direction of strong coupling vanish in highly anisotropic elliptic problems with homogeneous Neumann boundary data, a condition referred to as transverse equilibrium (TVE). To obtain a monotone coarse operator for heterogeneous problems the local elliptic problems used to determine the transmissibilities must be able to reach TVE as well. This can be achieved by solving two linear local problems with homogeneous Neumann boundary conditions and constructing a third bilinear local problem with Dirichlet boundary data taken from the linear local problems. Linear combination of these local problems gives the MSFV basis functions but with hybrid boundary conditions that cannot be enforced directly. The resulting compact multiscale finite volume (CMSFV) method with hybrid local boundary conditions is compared numerically to the original MSFV method. For isotropic problems both methods have comparable accuracy, but the CMSFV method is robust for highly anisotropic problems where the original MSFV method leads to unphysical oscillations in the coarse solution and recirculations in the reconstructed velocity field.

Multiscale finite-volume method for compressible multiphase flow in porous media The Multiscale Finite-Volume (MSFV) method has been recently developed and tested for multiphase-flow problems with simplified physics (i.e. incompressible flow without gravity and capillary effects) and proved robust, accurate and efficient. However, applications to practical problems necessitate extensions that enable the method to deal with more complex processes. In this paper we present a modified version of the MSFV algorithm that provides a suitable and natural framework to include additional physics. The algorithm consists of four main steps: computation of the local basis functions, which are used to extract the coarse-scale effective transmissibilities; solution of the coarse-scale pressure equation; reconstruction of conservative fine-scale fluxes; and solution of the transport equations. Within this framework, we develop a MSFV method for compressible multiphase flow. The basic idea is to compute the basis functions as in the case of incompressible flow such that they remain independent of the pressure. The effects of compressibility are taken into account in the solution of the coarse-scale pressure equation and, if necessary, in the reconstruction of the fine-scale fluxes. We consider three models with an increasing level of complexity in the flux reconstruction and test them for highly compressible flows (tracer transport in gas flow, imbibition and drainage of partially saturated reservoirs, depletion of gas-water reservoirs, and flooding of oil-gas reservoirs). We demonstrate that the MSFV method provides accurate solutions for compressible multiphase flow problems. Whereas slightly compressible flows can be treated with a very simple model, a more sophisticate flux reconstruction is needed to obtain accurate fine-scale saturation fields in highly compressible flows.

A mixed multiscale finite element method for elliptic problems with oscillating coefficients The recently introduced multiscale finite element method for solving elliptic equations with oscillating coefficients is designed to capture the large-scale structure of the solutions without resolving all the fine-scale structures. Motivated by the numerical simulation of flow transport in highly heterogeneous porous media, we propose a mixed multiscale finite element method with an over-sampling technique for solving second order elliptic equations with rapidly oscillating coefficients. The multiscale finite element bases are constructed by locally solving Neumann boundary value problems. We provide a detailed convergence analysis of the method under the assumption that the oscillating coefficients are locally periodic. While such a simplifying assumption is not required by our method, it allows us to use homogenization theory to obtain the asymptotic structure of the solutions. Numerical experiments are carried out for flow transport in a porous medium with a random log-normal relative permeability to demonstrate the efficiency and accuracy of the proposed method.

A hierarchical fracture model for the iterative multiscale finite volume method An iterative multiscale finite volume (i-MSFV) method is devised for the simulation of multiphase flow in fractured porous media in the context of a hierarchical fracture modeling framework. Motivated by the small pressure change inside highly conductive fractures, the fully coupled system is split into smaller systems, which are then sequentially solved. This splitting technique results in only one additional degree of freedom for each connected fracture network appearing in the matrix system. It can be interpreted as an agglomeration of highly connected cells; similar as in algebraic multigrid methods. For the solution of the resulting algebraic system, an i-MSFV method is introduced. In addition to the local basis and correction functions, which were previously developed in this framework, local fracture functions are introduced to accurately capture the fractures at the coarse scale. In this multiscale approach there exists one fracture function per network and local domain, and in the coarse scale problem there appears only one additional degree of freedom per connected fracture network. Numerical results are presented for validation and verification of this new iterative multiscale approach for fractured porous media, and to investigate its computational efficiency. Finally, it is demonstrated that the new method is an effective multiscale approach for simulations of realistic multiphase flows in fractured heterogeneous porous media.

Empirical Analysis of Predictive Algorithms for Collaborative Filtering Collaborative filtering or recommender systemsuse a database about user preferences topredict additional topics or products a newuser might like. In this paper we describe several algorithms designed for this task, including techniques based on correlation coefficients,vector-based similarity calculations,and statistical Bayesian methods. We comparethe predictive accuracy of the various methods in a set of representative problemdomains. We use two basic classes of evaluation...

Temporal data base management Reasoning about time typically involves drawing conclusions on the basis of incomplete information. Uncertainty arises in the form of ignorance, indeterminacy, and indecision. Despite the lack of complete information, a problem solver is continually forced to make predictions in order to pursue hypotheses and plan for the future. Such predictions are frequently contravened by subsequent evidence. This paper presents a computational approach to temporal reasoning that directly confronts these issues. The approach centers around techniques for managing a data base of assertions corresponding to the occurrence of events and the persistence of their effects over time. The resulting computational framework performs the temporal analog of (static) reason maintenance by keeping track of dependency information involving assumptions about the truth of facts spanning various intervals of time. The system described in this paper extends classical predicate-calculus data bases, such as those used by PROLOG, to deal with time in an efficient and natural manner.

Rational Kernels: Theory and Algorithms Many classification algorithms were originally designed for fixed-size vectors. Recent applications in text and speech processing and computational biology require however the analysis of variable-length sequences and more generally weighted automata. An approach widely used in statistical learning techniques such as Support Vector Machines (SVMs) is that of kernel methods, due to their computational efficiency in high-dimensional feature spaces. We introduce a general family of kernels based on weighted transducers or rational relations, rational kernels , that extend kernel methods to the analysis of variable-length sequences or more generally weighted automata. We show that rational kernels can be computed efficiently using a general algorithm of composition of weighted transducers and a general single-source shortest-distance algorithm. Not all rational kernels are positive definite and symmetric (PDS), or equivalently verify the Mercer condition, a condition that guarantees the convergence of training for discriminant classification algorithms such as SVMs. We present several theoretical results related to PDS rational kernels. We show that under some general conditions these kernels are closed under sum, product, or Kleene-closure and give a general method for constructing a PDS rational kernel from an arbitrary transducer defined on some non-idempotent semirings. We give the proof of several characterization results that can be used to guide the design of PDS rational kernels. We also show that some commonly used string kernels or similarity measures such as the edit-distance, the convolution kernels of Haussler, and some string kernels used in the context of computational biology are specific instances of rational kernels. Our results include the proof that the edit-distance over a non-trivial alphabet is not negative definite, which, to the best of our knowledge, was never stated or proved before. Rational kernels can be combined with SVMs to form efficient and powerful techniques for a variety of classification tasks in text and speech processing, or computational biology. We describe examples of general families of PDS rational kernels that are useful in many of these applications and report the result of our experiments illustrating the use of rational kernels in several difficult large-vocabulary spoken-dialog classification tasks based on deployed spoken-dialog systems. Our results show that rational kernels are easy to design and implement and lead to substantial improvements of the classification accuracy.

The Eden System: A Technical Review The Eden project is a five year experiment in designing, building, and using an "integrated distributed" computing system. We are attempting to combine the benefits of integration and distribution by supporting an object based style of programming on top of a node machine/local network hardware base. Our experimental hypothesis is that such an architecture will provide an environment conducive to building distributed applications.

Actions and specificity A solution to the problem of speciflcity in a resource{oriented deductive approach to actions and change is presented. Speciflcity originates in the problem of overloading methods in object oriented frameworks but can be observed in general applications of actions and change in logic. We give a uniform solution to the problem of speciflcity culminating in a completed equational logic program with an equational theory. We show the soundness and completeness of SLDENF{resolution, ie. SLD{resolution augmented by negation{as{failure and by an equational theory, wrt the completed program. Finally, the expressiveness of our approach for performing general reasoning about actions, change, and causality is demonstrated.

Data cache management using frequency-based replacement We propose a new frequency-based replacement algorithm for managing caches used for disk blocks by a file system, database management system, or disk control unit, which we refer to here as data caches. Previously, LRU replacement has usually been used for such caches. We describe a replacement algorithm based on the concept of maintaining reference counts in which locality has been “factored out”. In this algorithm replacement choices are made using a combination of reference frequency and block age. Simulation results based on traces of file system and I/O activity from actual systems show that this algorithm can offer up to 34% performance improvement over LRU replacement, where the improvement is expressed as the fraction of the performance gain achieved between LRU replacement and the theoretically optimal policy in which the reference string must be known in advance. Furthermore, the implementation complexity and efficiency of this algorithm is comparable to one using LRU replacement.

Improving the tolerance of multilayer perceptrons by minimizing the statistical sensitivity to weight deviations This paper proposes a version of the backpropagation algorithm which increases the tolerance of a feedforward neural network against deviations in the weight values. These changes can originate either when the neural network is mapped on a given VLSI circuit where the precision and/or weight matching are low, or by physical defects affecting the neural circuits. The modified backpropagation algorithm we propose uses the statistical sensitivity of the network to changes in the weights as a quantitative measure of network tolerance and attempts to reduce this statistical sensitivity while keeping the figures for the usual training performance (in errors and time) similar to those obtained with the usual backpropagation algorithm.

Exploring Sequence Alignment Algorithms On Fpga-Based Heterogeneous Architectures With the rapid development of DNA sequencer, the rate of data generation is rapidly outpacing the rate at which it can be computationally processed. Traditional sequence alignment based on PC cannot fulfill the increasing demand. Accelerating the algorithm using FPGA provides the better performance compared to the other platforms. This paper will explain and classify the current sequence alignment algorithms. In addition, we analyze the different types of sequence alignment algorithms and present the taxonomy of FPGA-based sequence alignment implementations. This work will conclude the current solutions and provide a reference to further accelerating sequence alignment on a FPGA-based heterogeneous architecture.

Practical prefetching techniques for parallel file systems Improvements in the processing speed of multiprocessors are outpacing improvements in the speed of disk hardware. Parallel disk I/O subsystems have been proposed as one way to close the gap between processor and disk speeds. In a previous paper the authors showed that prefetching and caching have the potential to deliver the performance benefits of parallel file systems to parallel applications. They describe experiments with practical prefetching policies, and show that prefetching can be implemented efficiently even for the more complex parallel file access patterns. They also test the ability of these policies across a range of architectural parameters. (see IEEE Trans. on Parallel and Distributed Systems, vol.1, no.2, p.218-30, 1990)

I/O-Conscious Volume Rendering Most existing volume rendering algorithms assume that data sets are memory-resident and thus ignore the performance overhead of disk I/O. While this assumption may be true for high-performance graphics machines, it does not hold for most desktop personal workstations. To minimize the end-to-end volume rendering time, this work re-examines implementation strategies of the ray casting algorithm, taking into account both computation and I/O overheads. Specifically, we developed a data-driven execution model for ray casting that achieves the maximum overlap between rendering computation and disk I/O. Together with other performance optimizations, on a 300-MHz Pentium-II machine, without directional shading, our implementation is able to render a 128x128 greyscale image from a 128x128x128 data set with an average end-to-end delay of 1 second, which is very close to the memory-resident rendering time. With a little modification, this work can also be extended to do out-of-core visualization as well.

High performance support of parallel virtual file system (PVFS2) over Quadrics Parallel I/O needs to keep pace with the demand of high performance computing applications on systems with ever-increasing speed. Exploiting high-end interconnect technologies to reduce the network access cost and scale the aggregated bandwidth is one of the ways to increase the performance of storage systems. In this paper, we explore the challenges of supporting parallel file system with modern features of Quadrics, including user-level communication and RDMA operations. We design and implement a Quadrics-capable version of a parallel file system (PVFS2). Our design overcomes the challenges imposed by Quadrics static communication model to dynamic client/server architectures. Quadrics QDMA and RDMA mechanisms are integrated and optimized for high performance data communication. Zero-copy PVFS2 list IO is achieved with a Single Event Associated MUltiple RDMA (SEAMUR) mechanism. Experimental results indicate that the performance of PVFS2, with Quadrics user-level protocols and RDMA operations, is significantly improved in terms of both data transfer and management operations. With four IO server nodes, our implementation improves PVFS2 aggregated read bandwidth by up to 140% compared to PVFS2 over TCP on top of Quadrics IP implementation. Moreover, it delivers significant performance improvement to application benchmarks such as mpi-tile-io [24] and BTIO [26]. To the best of our knowledge, this is the first work in the literature to report the design of a high performance parallel file system over Quadrics user-level communication protocols.

DualFS: a new journaling file system without meta-data duplication In this paper we introduce DualFS, a new high performance journaling file system that puts data and meta-data on different devices (usually, two partitions on the same disk or on different disks), and manages them in very different ways. Unlike other journaling file systems, DualFS has only one copy of every meta-data block. This copy is in the meta-data device, a log which is used by DualFS both to read and to write meta-data blocks. By avoiding a time-expensive extra copy of meta-data blocks, DualFS can achieve a good performance as compared to other journaling file systems. Indeed, we have implemented a DualFS prototype, which has been evaluated with microbenchmarks and macrobenchmarks, and we have found that DualFS greatly reduces the total I/O time taken by the file system in most cases (up to 97%), whereas it slightly increases the total I/O time only in a few and limited cases.

Informed prefetching of collective input/output requests

A Decoupled Architecture for Application-Specific File Prefetching Data-intensive applications such as multimedia and data mining programs may exhibit sophisticated access patterns that are difficult to predict from past reference history and are different from one application to, another. This paper presents the design, implementation, and evaluation of an automatic application-specific file prefetching (AASFP) mechanism that is designed to improve the disk I/O performance of application programs with such complicated access patterns. The key idea of AASFP is to convert an application into two threads: a computation thread, which is the original program containing both computation and disk I/O, and a prefetch thread, which contains all the instructions in the original program that are related to disk accesses. At run time, the prefetch thread is scheduled to run sufficiently far ahead of the computation thread, so that disk blocks can be prefetched and put in the file buffer cache before the computation thread needs them. Through a source-to-source translator, the conversion of a given application into two such threads is made completely automatic. Measurements on an initial AASFP prototype under Linux show that it provides as much as 54% overall performance improvement for a volume visualization application.

Implementation and performance of application-controlled file caching Traditional file system implementations do not allow applications to control file caching replacement decisions. We have implemented two-level replacement, a scheme that allows applications to control their own cache replacement, while letting the kernel control the allocation of cache space among processes. We designed an interface to let applications exert control on replacement via a set of directives to the kernel. This is effective and requires low overhead. We demonstrate that for applications that do not perform well under traditional caching policies, the combination of good application-chosen replacement strategies, and our kernel allocation policy LRU-SP, can reduce the number of block I/Os by up to 80%, and can reduce the elapsed time by up to 45%. We also show that LRU-SP is crucial to the performance improvement for multiple concurrent applications: LRU-SP fairly distributes cache blocks and offers protection against foolish applications.

Constant time permutation: an efficient block allocation strategy for variable-bit-rate continuous media data To provide high accessibility of continuous-media (CM) data, CM servers generally stripe data across multiple disks. Currently, the most widely used striping scheme for CM data is round-robin permutation (RRP). Unfortunately, when RRP is applied to variable-bit-rate (VBR) CM data, load imbalance across multiple disks occurs, thereby reducing overall system performance. In this paper, the performance of a VBR CM server with RRP is analyzed. In addition, we propose an efficient striping scheme called constant time permutation (CTP), which takes the VBR characteristic into account and obtains a more balanced load than RRP. Analytic models of both RRP and CTP are presented, and the models are verified via trace-driven simulations. Analysis and simulation results show that CTP can substantially increase the number of clients supported, though it might introduce a few seconds/minutes of initial delay.

Application-controlled physical memory using external page-cache management Next generation computer systems will have gigabytes of physical memory and processors in the 100 MIPS range or higher. Contrary to some conjectures, this trend requires more sophisticated memory management support for memory-bound computations such as scientific simulations and systems such as large-scale database systems, even though memory management for most programs will be less of a concern. We describe the design, implementation and evaluation of a virtual memory system that provides application control of physical memory using external page-cache management. In this approach, a sophisticated application is able to monitor and control the amount of physical memory it has available for execution, the exact contents of this memory, and the scheduling and nature of page-in and page-out using the abstraction of a physical page cache provided by the kernel. We claim that this approach can significantly improve performance for many memory-bound applications while reducing kernel complexity, yet does not complicate other applications or reduce their performance.

An Analytic Treatment Of The Reliability And Performance Of Mirrored Disk Subsystems

IRON file systems Commodity file systems trust disks to either work or fail completely, yet modern disks exhibit more complex failure modes. We suggest a new fail-partial failure model for disks, which incorporates realistic localized faults such as latent sector errors and block corruption. We then develop and apply a novel failure-policy fingerprinting framework, to investigate how commodity file systems react to a range of more realistic disk failures. We classify their failure policies in a new taxonomy that measures their Internal RObustNess (IRON), which includes both failure detection and recovery techniques. We show that commodity file system failure policies are often inconsistent, sometimes buggy, and generally inadequate in their ability to recover from partial disk failures. Finally, we design, implement, and evaluate a prototype IRON file system, Linux ixt3, showing that techniques such as in-disk checksumming, replication, and parity greatly enhance file system robustness while incurring minimal time and space overheads.

Many-layered learning. We explore incremental assimilation of new knowledge by sequential learning. Of particular interest is how a network of many knowledge layers can be constructed in an on-line manner, such that the learned units represent building blocks of knowledge that serve to compress the overall representation and facilitate transfer. We motivate the need for many layers of knowledge, and we advocate sequential learning as an avenue for promoting the construction of layered knowledge structures. Finally, our novel STL algorithm demonstrates a method for simultaneously acquiring and organizing a collection of concepts and functions as a network from a stream of unstructured information.

Concurrent Constraint Programming with Process Mobility We propose an extension of concurrent constraint programming with primitives for process migration within a hierarchical network, and we study its semantics. To this purpose, we first investigate a "pure" paradigm for process migration, namely a paradigm where the only actions are those dealing with transmissions of processes. Our goal is to give a structural definition of the semantics of migration; namely, we want to describe the behaviour of the system, during the transmission of a process, in terms of the behaviour of the components. We achieve this goal by using a labeled transition system where the effects of sending a process, and requesting a process, are modeled by symmetric rules (similar to handshaking-rules for synchronous communication) between the two partner nodes in the network. Next, we extend our paradigm with the primitives of concurrent constraint programming, and we show how to enrich the semantics to cope with the notions of environment and constraint store. Finally, we show how the operational semantics can be used to define an interpreter for the basic calculus.

Learning A Lexical Simplifier Using Wikipedia In this paper we introduce a new lexical simplification approach. We extract over 30K candidate lexical simplifications by identifying aligned words in a sentence-aligned corpus of English Wikipedia with Simple English Wikipedia. To apply these rules, we learn a feature-based ranker using SVMnk trained on a set of labeled simplifications collected using Amazon's Mechanical Turk. Using human simplifications for evaluation, we achieve a precision of 76% with changes in 86% of the examples.

The 1-Versus-2 Queries Problem Revisited The 1-versus-2 queries problem, which has been extensively studied in computational complexity theory, asks in its generality whether every efficient algorithm that makes at most 2 queries to a Σ k p -complete language L k has an efficient simulation that makes at most 1 query to L k . We obtain solutions to this problem for hypotheses weaker than previously considered. We prove that: For each k≥2, $\mathrm{P}^{\Sigma^{p}_{k}[2]}_{tt}\subseteq \mathrm{ZPP}^{\Sigma^{p}_{k}[1]}\Rightarrow \mathrm{PH}=\Sigma^{p}_{k}$, and PttNP[2]⊆ZPPNP[1]⇒PH=S2p. Here, for any complexity class $\mathcal{C}$and integer j≥1, we define $\mathrm{ZPP}^{\mathcal{C}[j]}$to be the class of problems solvable by zero-error randomized algorithms that run in polynomial time, make at most j queries to $\mathcal{C}$, and succeed with probability at least 1/2+1/poly(⋅). This same definition of $\mathrm{ZPP}^{\mathcal{C}[j]}$, also considered in Cai and Chakaravarthy (J. Comb. Optim. 11(2):189–202, 2006), subsumes the class of problems solvable by randomized algorithms that always answer correctly in expected polynomial time and make at most j queries to $\mathcal{C}$. Hemaspaandra, Hemaspaandra, and Hempel (SIAM J. Comput. 28(2):383–393, 1998), for k2, and Buhrman and Fortnow (J. Comput. Syst. Sci. 59(2):182–194, 1999), for k=2, had obtained the same consequence as ours in (I) using the stronger hypothesis $\mathrm{P}^{\Sigma^{p}_{k}[2]}_{tt}\subseteq \mathrm{P}^{\Sigma^{p}_{k}[1]}$. Fortnow, Pavan, and Sengupta (J. Comput. Syst. Sci. 74(3):358–363, 2008) had obtained the same consequence as ours in (II) using the stronger hypothesis P tt NP[2]⊆PNP[1]. Our results may also be viewed as steps towards obtaining solutions to arguably the most general form of the 1-versus-2 queries problem: For any k≥1, whether $\mathrm{P}^{\Sigma^{p}_{k}[2]}_{tt}$can be simulated in $\mathrm{BPP}^{\Sigma^{p}_{k}[1]}$.

Saving queries with randomness In this paper, we investigate the power of randomness to save a query to an NP-complete set. We show that the P SAT ∥ [ k ] ≤ p m -complete language randomly reduces to a language in P SAT ∥ [ k − 1] with a one-sided error probability of 1/⌈ k /2⌉ or a two sided-error probability of 1/( k +1). Furthermore, we prove that these probability bounds are tight; i.e., they cannot be improved by 1/poly, unless PH collapses. We also obtain tight performance bounds for randomized reductions between nearby classes in the Boolean and bounded query hierarchies. These bounds provide probability thresholds for completeness under randomized reductions in these classes. Using these thresholds, we show that certain languages in the Boolean hierarchy which are not ≤ p m -complete in some relativized worlds, nevertheless inherit many of the hardness properties associated with the ≤ p m -complete languages. Finally, we explore the relationship between randomization and functions that are computable using bounded queries to SAT. For any function h ( n ) = O (log n ), we show that there is a function f computable using h ( n ) nonadaptive queries to SAT, which cannot be computed correctly with probability 1/2 + 1/poly by any randomized machine which makes less than h ( n ) adaptive queries to any oracle, unless PH collapses.

Banishing Robust Turing Completeness ABSTRACT This paper proves that \promise classes" are so fragilely structured that they do not robustly (i.e. with respect to all oracles) possess Turinghard sets even in classes far larger than themselves. In particular, this paper shows that FewP does not robustly possess Turing hard sets for UP \ coUP and IP \ coIP does not robustly possess Turing hard sets for ZPP. It follows that ZPP, R, coR, UP\coUP, UP, FewP\coFewP, FewP, and IP \ coIP do not robustly possess Turing complete sets. This both resolves open questions of whether promise classes lacking robust downward closure under Turing reductions (e.g., R, UP, FewP) might robustly have Turing complete sets, and extends the range of classes known not to robustly contain many-one complete sets. Keywords: Structural complexity theory; Polynomial-time reductions;

A Downward Collapse within the Polynomial Hierarchy Downward collapse (also known as upward separation) refers to cases where the equality of two larger classes implies the equality of two smaller classes. We provide an unqualified downward collapse result completely within the polynomial hierarchy. In particular, we prove that, for k 2, if ${\rm P}^{\Sigma^p_k[1]} = {\rm P}^{\Sigma^p_k[2]}$ then $\Sigma^p_k = \Pi^p_k = {\rm PH}$. We extend this to obtain a more general downward collapse result.

Some connections between bounded query classes and non-uniform complexity It is shown that if there is a polynomial-time algorithm that tests k(n)=O(log n) points for membership in a set A by making only k(n)-1 adaptive queries to an oracle set X, then A belongs to NP/poly intersection co-NP/poly (if k(n)=O(1) then A belong to P/poly). In particular, k(n)=O(log n) queries to an NP -complete set (k(n)=O(1) queries to an NP-hard set) are more powerful than k(n)-1 queries, unless the polynomial hierarchy collapses. Similarly, if there is a small circuit that tests k(n) points for membership in A by making only k(n)-1 adaptive queries to a set X, then there is a correspondingly small circuit that decides membership in A without an oracle. An investigation is conducted of the quantitatively stronger assumption that there is a polynomial-time algorithm that tests 2k strings for membership in A by making only k queries to an oracle X, and qualitatively stronger conclusions about the structure of A are derived: A cannot be self-reducible unless A∈P, and A cannot be NP-hard unless P=NP. Similar results hold for counting classes. In addition, relationships between bounded-query computations, lowness, and the p-degrees are investigated

Simultaneous Strong Separations of Probabilistic and Unambiguous Complexity Classes We study the relationship between probabilistic and unambiguous computation, and provide strong relativized evidence that they are incomparable. In particular, we display a relativized world in which the complexity classes embodying these paradigms of computation are mutually immune. We answer questions formulated in|and extend the line of research opened by|Geske and Grollman (15) and Balcazar and Russo (3).

Two remarks on the power of counting The relationship between the polynomial hierarchy and Valiant's class #P is at present unknown. We show that some low portions of the polynomial hierarchy, namely deterministic polynomial algorithms using an NP oracle at most a logarithmic number of times, can be simulated by one #P computation. We also show that the class of problems solvable by polynomial-time nondeterministic Turing machines which accept whenever there is an odd number of accepting computations is idempotent, that is, closed under usage of oracles from the same class.

On the facial structure of set packing polyhedra In this paper we address ourselves to identifying facets of the set packing polyhedron, i.e., of the convex hull of integer solutions to the set covering problem with equality constraints and/or constraints of the form “?”. This is done by using the equivalent node-packing problem derived from the intersection graph associated with the problem under consideration. First, we show that the cliques of the intersection graph provide a first set of facets for the polyhedron in question. Second, it is shown that the cycles without chords of odd length of the intersection graph give rise to a further set of facets. A rather strong geometric property of this set of facets is exhibited.

LIBSVM: A library for support vector machines LIBSVM is a library for Support Vector Machines (SVMs). We have been actively developing this package since the year 2000. The goal is to help users to easily apply SVM to their applications. LIBSVM has gained wide popularity in machine learning and many other areas. In this article, we present all implementation details of LIBSVM. Issues such as solving SVM optimization problems theoretical convergence multiclass classification probability estimates and parameter selection are discussed in detail.

An empirical evaluation of deep architectures on problems with many factors of variation Recently, several learning algorithms relying on models with deep architectures have been proposed. Though they have demonstrated impressive performance, to date, they have only been evaluated on relatively simple problems such as digit recognition in a controlled environment, for which many machine learning algorithms already report reasonable results. Here, we present a series of experiments which indicate that these models show promise in solving harder learning problems that exhibit many factors of variation. These models are compared with well-established algorithms such as Support Vector Machines and single hidden-layer feed-forward neural networks.

Near-Optimal Plans, Tractability, and Reactivity Many planning problems have recently beenshown to be inherently intractable. For example,finding the shortest plan in the blocksworlddomain is NP-hard, and so is planningin even some of the most limited STRIPSstyleplanning formalisms. We explore thequestion as to what extent these negative resultscan be attributed to the insistence onfinding plans of minimal length.Using recent results form the theory of combinatorialoptimization, we show that fordomain-independent planning, one...

The control of reasoning in resource-bounded agents Autonomous agents are systems capable of autonomous decision-making in real-time environments. Computation is a valuable resource for such decision-making, and yet the amount of computation that an autonomous agent may carry out will be limited. It follows that an agent must be equipped with a mechanism that enables it to make the best possible use of the computational resources at its disposal. In this paper we review three approaches to the control of computation in resource-bounded agents. In addition to a detailed description of each framework, this paper compares and contrasts the approaches, and lists the advantages and disadvantages of each.

On the Complexity of Plan Adaptation by Derivational Analogy in a Universal Classical Planning Framework In this paper we present an algorithm called DerUCP, which can be regarded as a general model for plan adaptation using Derivational Analogy. Using DerUCP, we show that previous results on the complexity of plan adaptation do not apply to Derivational Analogy. We also show that Derivational Analogy can potentially produce exponential reductions in the size of the search space generated by a planning system.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Computational Modeling of Mammalian Promoters - (Invited Keynote Talk).

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Scene parsing using inference Embedded Deep Networks. Effective features and graphical model are two key points for realizing high performance scene parsing. Recently, Convolutional Neural Networks (CNNs) have shown great ability of learning features and attained remarkable performance. However, most researches use CNNs and graphical model separately, and do not exploit full advantages of both methods. In order to achieve better performance, this work aims to design a novel neural network architecture called Inference Embedded Deep Networks (IEDNs), which incorporates a novel designed inference layer based on graphical model. Through the IEDNs, the network can learn hybrid features, the advantages of which are that they not only provide a powerful representation capturing hierarchical information, but also encapsulate spatial relationship information among adjacent objects. We apply the proposed networks to scene labeling, and several experiments are conducted on SIFT Flow and PASCAL VOC Dataset. The results demonstrate that the proposed IEDNs can achieve better performance. HighlightsWe design a novel structure of networks considering CRFs model as one type layer of deep neural networks.CRF is regarded as a layer of the network, therefore, the structural learning can be conducted explicitly.A novel feature encoding spatial relationship between objects in images is proposed.Feature fusing is adopted to learn intrinsic non-linear relationships between hierarchical and spatial features.

Learning Deep Architectures for AI Theoretical results suggest that in order to learn the kind of com-plicated functions that can represent high-level abstractions (e.g., invision, language, and other AI-level tasks), one may needdeep architec-tures. Deep architectures are composed of multiple levels of non-linearoperations, such as in neural nets with many hidden layers or in com-plicated propositional formulae re-using many sub-formulae. Searchingthe parameter space of deep architectures is a difficult task, but learningalgorithms such as those for Deep Belief Networks have recently beenproposed to tackle this problem with notable success, beating the state-of-the-art in certain areas. This monograph discusses the motivationsand principles regarding learning algorithms for deep architectures, inparticular those exploiting as building blocks unsupervised learning ofsingle-layer models such as Restricted Boltzmann Machines, used toconstruct deeper models such as Deep Belief Networks.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Logic programs with classical negation

The well-founded semantics for general logic programs A general logic program (abbreviated to “program” hereafter) is a set of roles that have both positive and negative subgoals. It is common to view a deductive database as a general logic program consisting of rules (IDB) slttmg above elementary relations (EDB, facts). It is desirable to associate one Herbrand model with a program and think of that model as the “meaning of the program, ” or Its“declarative semantics. ” Ideally, queries directed to the program would be answered in accordance with this model. Recent research indicates that some programs do not have a “satisfactory” total model; for such programs, the question of an appropriate partial model arises. Unfounded sets and well-founded partial models are introduced and the well-founded semantics of a program are defined to be its well-founded partial model. If the well-founded partial model is m fact a total model. it is called the well-founded model. It n shown that the class of programs possessing a total well-founded model properly includes previously studied classes of “stratified” and “locally stratified” programs,The method in this paper is also compared with other proposals in the literature, including Clark’s“program completion, ” Fitting’s and Kunen’s 3-vahred interpretations of it, and the “stable models”of Gelfond and Lifschitz.

Solving Advanced Reasoning Tasks Using Quantified Boolean Formulas We consider the compilation of different reasoning tasks into the evaluation problem of quantified boolean formulas (QBFs) as an approach to develop prototype reasoning sys- tems useful for, e.g., experimental purposes. Such a method is a natural generalization of a similar technique applied to NP-problems and has been recently proposed by other re- searchers. More specifically, we present translations of sev- eral well-known reasoning tasks from the area of nonmono- tonic reasoning into QBFs, and compare their implementa- tion in the prototype system QUIP with established NMR- provers. The results show reasonable performance, and docu- ment that the QBF approach is an attractive tool for rapid pro- totyping of experimental knowledge-representation systems.

Object Recognition from Local Scale-Invariant Features An object recognition system has been developed that uses a new class of local image features. The features are invariant to image scaling, translation, and rotation, and partially invariant to illumination changes and affine or 3D projection.These features share similar properties with neurons in inferior temporal cortex that are used for object recognition in primate vision. Features are efficiently detected through a staged filtering approach that identifies stable points in scale space. Image keys are created that allow for local geometric deformations by representing blurred image gradients in multiple orientation planes and at multiple scales.The keys are used as input to a nearest-neighbor indexing method that identifies candidate object matches. Final verification of each match is achieved by finding a low-residual least-squares solution for the unknown model parameters. Experimental results show that robust object recognition can be achieved in cluttered partially-occluded images with a computation time of under 2 seconds.

Support-Vector Networks The support-vector network is a new learning machine for two-group classification problems. The machine conceptually implements the following idea: input vectors are non-linearly mapped to a very high-dimension feature space. In this feature space a linear decision surface is constructed. Special properties of the decision surface ensures high generalization ability of the learning machine. The idea behind the support-vector network was previously implemented for the restricted case where the training data can be separated without errors. We here extend this result to non-separable training data.High generalization ability of support-vector networks utilizing polynomial input transformations is demonstrated. We also compare the performance of the support-vector network to various classical learning algorithms that all took part in a benchmark study of Optical Character Recognition.

Improving the I/O Performance of Real-Time Database Systems with Multiple-Disk Storage Structures

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Learning Topic Representation For Smt With Neural Networks Statistical Machine Translation (SMT) usually utilizes contextual information to disambiguate translation candidates. However, it is often limited to contexts within sentence boundaries, hence broader topical information cannot be leveraged. In this paper, we propose a novel approach to learning topic representation for parallel data using a neural network architecture, where abundant topical contexts are embedded via topic relevant monolingual data. By associating each translation rule with the topic representation, topic relevant rules are selected according to the distributional similarity with the source text during SMT decoding. Experimental results show that our method significantly improves translation accuracy in the NIST Chinese-to-English translation task compared to a state-of-the-art baseline.

New Directions and New Challenges in Algorithm Design and Complexity, Parameterized The goals of this survey are to: (1) Motivate the basic notions of parameterized complexity and give some examples to introduce the toolkits of FPT and W-hardness as concretely as possible for those who are new to these ideas. (2) Describe some new research directions, new techniques and challenging open problems in this area.

Parameterized complexity for the database theorist

The Parameterized Complexity of Counting Problems We develop a parameterized complexity theory for counting problems. As the basis of this theory, we introduce a hierarchy of parameterized counting complexity classes #W$[t]$, for $t\ge 1$, that corresponds to Downey and Fellows's W-hierarchy [R. G. Downey and M. R. Fellows, Parameterized Complexity, Springer-Verlag, New York, 1999] and we show that a few central W-completeness results for decision problems translate to \#W-completeness results for the corresponding counting problems. Counting complexity gets interesting with problems whose decision version is tractable, but whose counting version is hard. Our main result states that counting cycles and paths of length k in both directed and undirected graphs, parameterized by k, is #W$[1]-complete. This makes it highly unlikely that these problems are fixed-parameter tractable, even though their decision versions are fixed-parameter tractable. More explicitly, our result shows that most likely there is no $f(k) \cdot n^c$-algorithm for counting cycles or paths of length k in a graph of size n for any computable function $f: \mathbb{N} \to \mathbb{N}$ and constant c, even though there is a $2^{O(k)} \cdot n^{2.376}$ algorithm for finding a cycle or path of length k [N. Alon, R. Yuster, and U. Zwick, J. ACM, 42 (1995), pp. 844--856].

Formal methods for the validation of automotive product configuration data Constraint-based reasoning is often used to represent and find solutions to configuration problems. In the field of constraint satisfaction, the major focus has been on finding solutions to difficult problems. However, many real-life configuration problems, ...

Fixed-parameter complexity in AI and nonmonotonic reasoning Many relevant intractable problems become tractable if some problem parameter is fixed. However, various problems exhibit very different computational properties, depending on how the runtime required for solving them is related to the fixed parameter chosen. The theory of parameterized complexity deals with such issues, and provides general techniques for identifying fixed-parameter tractable and fixed-parameter intractable problems. We study the parameterized complexity of various problems in AI and nonmonotonic reasoning. We show that a number of relevant parameterized problems in these areas are fixed-parameter tractable. Among these problems are constraint satisfaction problems with bounded treewidth and fixed domain, restricted forms of conjunctive database queries, restricted satisfiability problems, propositional logic programming under the stable model semantics where the parameter is the dimension of a feedback vertex set of the program's dependency graph, and circumscriptive inference from a positive k-CNF restricted to models of bounded size. We also show that circumscriptive inference from a general propositional theory, when the attention is restricted to models of bounded size, is fixed-parameter intractable and is actually complete for a novel fixed-parameter complexity class.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Evaluating collaborative filtering recommender systems Recommender systems have been evaluated in many, often incomparable, ways. In this article, we review the key decisions in evaluating collaborative filtering recommender systems: the user tasks being evaluated, the types of analysis and datasets being used, the ways in which prediction quality is measured, the evaluation of prediction attributes other than quality, and the user-based evaluation of the system as a whole. In addition to reviewing the evaluation strategies used by prior researchers, we present empirical results from the analysis of various accuracy metrics on one content domain where all the tested metrics collapsed roughly into three equivalence classes. Metrics within each equivalency class were strongly correlated, while metrics from different equivalency classes were uncorrelated.

Practical Issues in Temporal Difference Learning This paper examines whether temporal difference methods for training connectionist networks, such as Sutton's TD(λ) algorithm, can be successfully applied to complex real-world problems. A number of important practical issues are identified and discussed from a general theoretical perspective. These practical issues are then examined in the context of a case study in which TD(λ) is applied to learning the game of backgammon from the outcome of self-play. This is apparently the first application of this algorithm to a complex non-trivial task. It is found that, with zero knowledge built in, the network is able to learn from scratch to play the entire game at a fairly strong intermediate level of performance, which is clearly better than conventional commercial programs, and which in fact surpasses comparable networks trained on a massive human expert data set. This indicates that TD learning may work better in practice than one would expect based on current theory, and it suggests that further analysis of TD methods, as well as applications in other complex domains, may be worth investigating.

The Boolean hierarchy: hardware over NP In this paper, we study the complexity of sets formed by boolean operations $(\bigcup, \bigcap,$ and complementation) on NP sets. These are the sets accepted by trees of hardware with NP predicates as leaves, and together form the boolean hierarchy. We present many results about the boolean hierarchy: separation and immunity results, complete languages, upward separations, connections to sparse oracles for NP, and structural asymmetries between complementary classes. Some results present new ideas and techniques. Others put previous results about NP and $D^{P}$ in a richer perspective. Throughout, we emphasize the structure of the boolean hierarchy and its relations with more common classes.

A Stable Distributed Scheduling Algorithm

Normal forms for answer sets programming Normal forms for logic programs under stable/answer set semantics are introduced. We argue that these forms can simplify the study of program properties, mainly consistency. The first normal form, called the kernel of the program, is useful for studying existence and number of answer sets. A kernel program is composed of the atoms which are undefined in the Well-founded semantics, which are those that directly affect the existence of answer sets. The body of rules is composed of negative literals only. Thus, the kernel form tends to be significantly more compact than other formulations. Also, it is possible to check consistency of kernel programs in terms of colorings of the Extended Dependency Graph program representation which we previously developed. The second normal form is called 3-kernel. A 3-kernel program is composed of the atoms which are undefined in the Well-founded semantics. Rules in 3-kernel programs have at most two conditions, and each rule either belongs to a cycle, or defines a connection between cycles. 3-kernel programs may have positive conditions. The 3-kernel normal form is very useful for the static analysis of program consistency, i.e. the syntactic characterization of existence of answer sets. This result can be obtained thanks to a novel graph-like representation of programs, called Cycle Graph which presented in the companion article Costantini (2004b).

A cost-benefit scheme for high performance predictive prefetching

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Adaptive Parallel Tempering for Stochastic Maximum Likelihood Learning of RBMs Restricted Boltzmann Machines (RBM) have attracted a lot of attention of late, as one the principle building blocks of deep networks. Training RBMs remains problematic however, because of the intractibility of their partition function. The maximum likelihood gradient requires a very robust sampler which can accurately sample from the model despite the loss of ergodicity often incurred during learning. While using Parallel Tempering in the negative phase of Stochastic Maximum Likelihood (SML-PT) helps address the issue, it imposes a trade-off between computational complexity and high ergodicity, and requires careful hand-tuning of the temperatures. In this paper, we show that this trade-off is unnecessary. The choice of optimal temperatures can be automated by minimizing average return time (a concept first proposed by [Katzgraber et al., 2006]) while chains can be spawned dynamically, as needed, thus minimizing the computational overhead. We show on a synthetic dataset, that this results in better likelihood scores.

How to Center Binary Restricted Boltzmann Machines.

Training restricted boltzmann machines with multi-tempering: harnessing parallelization Restricted Boltzmann Machines (RBM's) are unsupervised probabilistic neural networks that can be stacked to form Deep Belief Networks. Given the recent popularity of RBM's and the increasing availability of parallel computing architectures, it becomes interesting to investigate learning algorithms for RBM's that benefit from parallel computations. In this paper, we look at two extensions of the parallel tempering algorithm, which is a Markov Chain Monte Carlo method to approximate the likelihood gradient. The first extension is directed at a more effective exchange of information among the parallel sampling chains. The second extension estimates gradients by averaging over chains from different temperatures. We investigate the efficiency of the proposed methods and demonstrate their usefulness on the MNIST dataset. Especially the weighted averaging seems to benefit Maximum Likelihood learning.

Training RBMs based on the signs of the CD approximation of the log-likelihood derivatives.

Enhanced Gradient and Adaptive Learning Rate for Training Restricted Boltzmann Machines.

The Neural Autoregressive Distribution Estimator

Convolutional deep belief networks for scalable unsupervised learning of hierarchical representations There has been much interest in unsupervised learning of hierarchical generative models such as deep belief networks. Scaling such models to full-sized, high-dimensional images remains a difficult problem. To address this problem, we present the convolutional deep belief network, a hierarchical generative model which scales to realistic image sizes. This model is translation-invariant and supports efficient bottom-up and top-down probabilistic inference. Key to our approach is probabilistic max-pooling, a novel technique which shrinks the representations of higher layers in a probabilistically sound way. Our experiments show that the algorithm learns useful high-level visual features, such as object parts, from unlabeled images of objects and natural scenes. We demonstrate excellent performance on several visual recognition tasks and show that our model can perform hierarchical (bottom-up and top-down) inference over full-sized images.

Detonation Classification from Acoustic Signature with the Restricted Boltzmann Machine We compare the recently proposed Discriminative Restricted Boltzmann Machine (DRBM) to the classical Support Vector Machine (SVM) on a challenging classification task consisting in identifying weapon classes from audio signals. The three weapon classes considered in this work (mortar, rocket, and rocket-propelled grenade), are difficult to reliably classify with standard techniques because they tend to have similar acoustic signatures. In addition, specificities of the data available in this study make it challenging to rigorously compare classifiers, and we address methodological issues arising from this situation. Experiments show good classification accuracy that could make these techniques suitable for fielding on autonomous devices. DRBMs appear to yield better accuracy than SVMs, and are less sensitive to the choice of signal preprocessing and model hyperparameters. This last property is especially appealing in such a task where the lack of data makes model validation difficult. (10Roughly speaking, the number of DOF in the regression residuals is computed as the number of observations in the training set minus the number of parameters that are part of the regression model. © 2012 Wiley Periodicals, Inc.)

Audio-based Music Classification with a Pretrained Convolutional Network.

A Hierarchical Model Of Shape And Appearance For Human Action Classification We present a novel model for human action categorization. A video sequence is represented as a collection of spatial and spatial-temporal features by extracting static and dynamic interest points. We propose a hierarchical model that can be characterized as a constellation of bags-of-features and that is able to combine both spatial and spatial-temporal features. Given a novel video sequence, the model is able to categorize human actions in a frame-by-frame basis. We test the model on a publicly available human action dataset (2) and show that our new method performs well on the classification task. We also conducted control experiments to show that the use of the proposed mixture of hierarchical models improves the classification performance over bag of feature models. An additional experiment shows that using both dynamic and static features provides a richer representation of human actions when compared to the use of a single feature type, as demonstrated by our evaluation in the classification task.

Parity logging disk arrays Parity-encoded redundant disk arrays provide highly reliable, cost-effective secondary storage with high performance for reads and large writes. Their performance on small writes, however, is much worse than mirrored disks—the traditional, highly reliable, but expensive organization for secondary storage. Unfortunately, small writes are a substantial portion of the I/O workload of many important, demanding applications such as on-line transaction processing. This paper presents parity logging, a novel solution to the small-write problem for redundant disk arrays. Parity logging applies journalling techniques to reduce substantially the cost of small writes. We provide detailed models of parity logging and competing schemes—mirroring, floating storage, and RAID level 5—and verify these models by simulation. Parity logging provides performance competitive with mirroring, but with capacity overhead close to the minimum offered by RAID level 5. Finally, parity logging can exploit data caching more effectively than all three alternative approaches.

Concurrent actions in the situation calculus We propose a representation of Concurrent actions; rather than invent a new formalism, we model them within the standard situation calculus by introducing the notions of global actions and primitive actions, whose relationship is analogous to' that between situations and fluents. The result is a framework in which situations and actions play quite symmetric roles. The rich structure of actions gives rise to' a new problem, which, due to' this symmetry between actions and situations, is analogous to' the traditional frame problem. In [Lin and Shoham 1991] we provided a solution to' the frame problem based on a formal adequacy criterion called "epistemological completeness." Here we show how to' solve the new problem based on the same adequacy criterion.

Comparing Different Prenexing Strategies for Quantified Boolean Formulas The majority of the currently available solvers for quantified Boolean formulas (QBFs) process input formulas only in prenex conjunctive normal form. However, the natural representation of practicably relevant problems in terms of QBFs usually results in formulas which are not in a specific normal form. Hence, in order to evaluate such QBFs with available solvers, suitable normal-form translations are required. In this paper, we report experimental results comparing different prenexing strategies on a class of structured benchmark problems. The problems under consideration encode the evaluation of nested counterfactuals over a propositional knowledge base, and span the entire polynomial hierarchy. The results show that different prenexing strategies influence the evaluation time in different ways across different solvers. In particular, some solvers are robust to the chosen strategies while others are not.

Plan aggregation for strong cyclic planning in nondeterministic domains. We describe a planning algorithm, NDP2, that finds strong-cyclic solutions to nondeterministic planning problems by using a classical planner to solve a sequence of classical planning problems. NDP2 is provably correct, and fixes several problems with prior work.

Neither a Bazaar nor a cathedral: The interplay between structure and agency in Wikipedia's role system AbstractRoles provide a key coordination mechanism in peer‐production. Whereas one stream in the literature has focused on the structural responsibilities associated with roles, another has stressed the emergent nature of work. To date, these streams have proceeded largely in parallel. In seeking to enhance our understanding of the tension between structure and agency in peer‐production, we investigated the interplay between structural and emergent roles. Our study explored the breadth of structural roles in Wikipedia (English version) and their linkage to various forms of activities. Our analyses show that despite the latitude in selecting their mode of participation, participants' structural and emergent roles are tightly coupled. Our discussion highlights that: (a) participants often stay close to the “production ground floor” despite the assignment into structural roles; and (b) there are typical modifications in activity patterns associated with role‐assignment, namely: functional specialization, multispecialization, defunctionalization, changes in communication patterns, management of identity, and role definition. We contribute to theory of coordination and roles, as well as provide some practical implications.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Learning Topic Representation For Smt With Neural Networks Statistical Machine Translation (SMT) usually utilizes contextual information to disambiguate translation candidates. However, it is often limited to contexts within sentence boundaries, hence broader topical information cannot be leveraged. In this paper, we propose a novel approach to learning topic representation for parallel data using a neural network architecture, where abundant topical contexts are embedded via topic relevant monolingual data. By associating each translation rule with the topic representation, topic relevant rules are selected according to the distributional similarity with the source text during SMT decoding. Experimental results show that our method significantly improves translation accuracy in the NIST Chinese-to-English translation task compared to a state-of-the-art baseline.

Quantifying privacy in multiagent planning Privacy is often cited as the main reason to adopt a multiagent approach for a certain problem. This also holds true for multiagent planning. Still, a metric to evaluate the privacy performance of planners is virtually non-existent. This makes it hard to compare different algorithms on their performance with regards to privacy. Moreover, it prevents multiagent planning methods from being designed specifically for this aspect. This paper introduces such a measure for privacy. It is based on Shannon's theory of information and revolves around counting the number of alternative plans that are consistent with information that is gained during, for example, a negotiation step, or the complete planning episode. To accurately obtain this measure, one should have intimate knowledge of the agent's domain. It is unlikely (although not impossible) that an opponent who learns some information on a target agent has this knowledge. Therefore, it is not meant to be used by an opponent to understand how much he has learned. Instead, the measure is aimed at agents who want to know how much privacy they have given up, or are about to give up, in the planning process. They can then use this to decide whether or not to engage in a proposed negotiation, or to limit the options they are willing to negotiate upon.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Deep Fisher Kernels -- End to End Learning of the Fisher Kernel GMM Parameters Fisher Kernels and Deep Learning were two developments with significant impact on large-scale object categorization in the last years. Both approaches were shown to achieve state-of-the-art results on large-scale object categorization datasets, such as ImageNet. Conceptually, however, they are perceived as very different and it is not uncommon for heated debates to spring up when advocates of both paradigms meet at conferences or workshops. In this work, we emphasize the similarities between both architectures rather than their differences and we argue that such a unified view allows us to transfer ideas from one domain to the other. As a concrete example we introduce a method for learning a support vector machine classifier with Fisher kernel at the same time as a task-specific data representation. We reinterpret the setting as a multi-layer feed forward network. Its final layer is the classifier, parameterized by a weight vector, and the two previous layers compute Fisher vectors, parameterized by the coefficients of a Gaussian mixture model. We introduce a gradient descent based learning algorithm that, in contrast to other feature learning techniques, is not just derived from intuition or biological analogy, but has a theoretical justification in the framework of statistical learning theory. Our experiments show that the new training procedure leads to significant improvements in classification accuracy while preserving the modularity and geometric interpretability of a support vector machine setup.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

A linear time algorithm for finding tree-decompositions of small treewidth In this paper, we give for constant k a linear-time algorithm that, given a graph G = (V, E), determines whether the treewidth of G is at most k and, if so, finds a tree-decomposition of G with treewidth at most k. A consequence is that every minor-closed class of graphs that does not contain all planar graphs has a linear-time recognition algorithm. Another consequence is that a similar result holds when we look instead for path-decompositions with pathwidth at mast some constant k.

Compendium of Parameterized Problems at Higher Levels of the Polynomial Hierarchy. We present a list of parameterized problems together with a complexity classification of whether they allow a fixed-parameter tractable reduction to SAT or not. These problems are parameterized versions of problems whose complexity lies at the second level of the Polynomial Hierarchy or higher.

Constraint satisfaction with bounded treewidth revisited The constraint satisfaction problem can be solved in polynomial time for instances where certain parameters (e.g., the treewidth of primal graphs) are bounded. However, there is a trade-off between generality and performance: larger bounds on the parameters yield worse time complexities. It is desirable to pay for more generality only by a constant factor in the running time, not by a larger degree of the polynomial. Algorithms with such a uniform polynomial time complexity are known as fixed-parameter algorithms. In this paper we determine whether or not fixed-parameter algorithms for constraint satisfaction exist, considering all possible combinations of the following parameters: the treewidth of primal graphs, the treewidth of dual graphs, the treewidth of incidence graphs, the domain size, the maximum arity of constraints, and the maximum size of overlaps of constraint scopes. The negative cases are subject to the complexity theoretic assumption FPT ≠ W[1] which is the parameterized analog to P ≠ NP. For the positive cases we provide an effective fixed-parameter algorithm which is based on dynamic programming on “nice” tree decompositions.

Algorithms for propositional model counting We present algorithms for the propositional model counting problem #SAT. The algorithms utilize tree decompositions of certain graphs associated with the given CNF formula; in particular we consider primal, dual, and incidence graphs. We describe the algorithms coherently for a direct comparison and with sufficient detail for making an actual implementation reasonably easy. We discuss several aspects of the algorithms including worst-case time and space requirements.

New Directions and New Challenges in Algorithm Design and Complexity, Parameterized The goals of this survey are to: (1) Motivate the basic notions of parameterized complexity and give some examples to introduce the toolkits of FPT and W-hardness as concretely as possible for those who are new to these ideas. (2) Describe some new research directions, new techniques and challenging open problems in this area.

The complexity of acyclic conjunctive queries This paper deals with the evaluation of acyclic Booleanconjunctive queries in relational databases. By well-known resultsof Yannakakis[1981], this problem is solvable in polynomial time;its precise complexity, however, has not been pinpointed so far. Weshow that the problem of evaluating acyclic Boolean conjunctivequeries is complete for LOGCFL, the class of decision problems thatare logspace-reducible to a context-free language. Since LOGCFL iscontained in AC1 and NC2, the evaluation problem of acyclic Booleanconjunctive queries is highly parallelizable. We present a paralleldatabase algorithm solving this problem with alogarithmic number ofparallel join operations. The algorithm is generalized to computingthe output of relevant classes of non-Boolean queries. We also showthat the acyclic versions of the following well-known database andAI problems are all LOGCFL-complete: The Query Output Tuple problemfor conjunctive queries, Conjunctive Query Containment, ClauseSubsumption, and Constraint Satisfaction. The LOGCFL-completenessresult is extended to the class of queries of bounded tree widthand to other relevant query classes which are more general than theacyclic queries.

The influence of k-dependence on the complexity of planning A planning problem is k-dependent if each action has at most k pre-conditions on variables unaffected by the action. This concept is of interest because k is a constant for all but a few of the current benchmark domains in planning, and is known to have implications for tractability. In this paper, we present an algorithm for solving planning problems in P(k), the class of k-dependent planning problems with binary variables and polytree causal graphs. We prove that our algorithm runs in polynomial time when k is a fixed constant. If, in addition, the causal graph has bounded depth, we show that plan generation is linear in the size of the input. Although these contributions are theoretical due to the limited scope of the class P(k), suitable reductions from more complex planning problems to P(k) could potentially give rise to fast domain-independent heuristics.

A theory of diagnosis from first principles Without Abstract

Analysis of search based algorithms for satisfiability of propositional and quantified boolean formulas arising from circuit state space diameter problems The sequential circuit state space diameter problem is an important problem in sequential verification. Bounded model checking is complete if the state space diameter of the system is known. By unrolling the transition relation, the sequential circuit state space diameter problem can be formulated as either a series of Boolean satisfiability (SAT) problems or an evaluation for satisfiability of a Quantified Boolean Formula (QBF). Thus far neither the SAT based technique that uses sophisticated SAT solvers, nor QBF evaluations for the various QBF formulations for this have fared well in practice. The poor performance of the QBF evaluations is blamed on the relative immaturity of QBF solvers, with hope that ongoing research in QBF solvers could lead to practical success here. Most existing QBF algorithms, such as those based on the DPLL SAT algorithm, are search based. We show that using search based QBF algorithms to calculate the state space diameter of sequential circuits with existing problem formulations is no better than using SAT to solve this problem. This result holds independent of the representation of the QBF formula. This result is important as it highlights the need to explore non-search based or hybrid of search and non-search based QBF algorithms for the sequential circuit state space diameter problem.

On the NP-ardness of blocks world Blocks world (cube world) has been one of the most popular model domains in artificial intelligence search and planning. The operation and effectiveness of alternative heuristic strategies, both basic and complex, can be observed easily in this domain. We show that finding an optimal solution is NP-hard in an important variant of the domain, and popular extensions. This enlarges the range of model domains whose complexity has been explored mathematically, and it demonstrates that the complexity of search in blocks world is on the same level as for sliding block problems, the traveling salesperson problem, binpacking problems, and the like. These results also support the practice of using blocks world as a tutorial search domain in courses on artificial intelligence, to reveal both the value and limitations of heuristic search when seeking optimal solutions.

Prediction is deduction but explanation is abduction This paper presents an approach to temporal reasoning in which prediction is deduction but explanation is abduction. It is argued that all causal laws should be expressed in the natural form effect if cause. Any given set of laws expressed in this way can be used for both forwards projection (prediction) and backwards projection (explanation), but abduction must be used for explanation whilst deduction is used for prediction. The approach described uses a shortened form of Kowalski and Sergot's Event Calculus and incorporates the assumption that properties known to hold must have explanations in terms of events. Using abduction to implement this assumption results in a form of default persistence which correctly handles problems which have troubled other formulations. A straightforward extension to SLD resolution is described which implements the abductive approach to explanation, and which complements the well-understood deductive methods for prediction.

Stacked Denoising Autoencoders: Learning Useful Representations in a Deep Network with a Local Denoising Criterion We explore an original strategy for building deep networks, based on stacking layers of denoising autoencoders which are trained locally to denoise corrupted versions of their inputs. The resulting algorithm is a straightforward variation on the stacking of ordinary autoencoders. It is however shown on a benchmark of classification problems to yield significantly lower classification error, thus bridging the performance gap with deep belief networks (DBN), and in several cases surpassing it. Higher level representations learnt in this purely unsupervised fashion also help boost the performance of subsequent SVM classifiers. Qualitative experiments show that, contrary to ordinary autoencoders, denoising autoencoders are able to learn Gabor-like edge detectors from natural image patches and larger stroke detectors from digit images. This work clearly establishes the value of using a denoising criterion as a tractable unsupervised objective to guide the learning of useful higher level representations.

Design And Implementation Of An Fpga-Based Core For Gapped Blast Sequence Alignment With The Two-Hit Method This paper presents the design and implementation of the first FPGA-based core for Gapped BLAST sequence alignment with the two-hit method, ever reported in the literature. Gapped BLAST with two hit is a heuristic biological sequence alignment algorithm which is very widely used in the Bioinformatics and Computational Biology world. The architecture of the core is parameterized in terms of sequence lengths, match scores, gap penalties and cut-off, and threshold values. It is composed of various blocks each of which performs one step of the algorithm in parallel. This results in high performance and efficient FPGA implementations, which easily outperform equivalent software implementations by one order of magnitude or more. Furthermore, the core was captured in an FPGA-platform-independent language, namely the Handel-C language, to which no specific resource inference or placement constraints were applied. Hence, the core can be ported to different FPGA families and architectures.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Disk cache—miss ratio analysis and design considerations The current trend of computer system technology is toward CPUs with rapidly increasing processing power and toward disk drives of rapidly increasing density, but with disk performance increasing very slowly if at all. The implication of these trends is that at some point the processing power of computer systems will be limited by the throughput of the input/output (I/O) system.A solution to this problem, which is described and evaluated in this paper, is disk cache. The idea is to buffer recently used portions of the disk address space in electronic storage. Empirically, it is shown that a large (e.g., 80-90 percent) fraction of all I/O requests are captured by a cache of an 8-Mbyte order-of-magnitude size for our workload sample. This paper considers a number of design parameters for such a cache (called cache disk or disk cache), including those that can be examined experimentally (cache location, cache size, migration algorithms, block sizes, etc.) and others (access time, bandwidth, multipathing, technology, consistency, error recovery, etc.) for which we have no relevant data or experiments. Consideration is given to both caches located in the I/O system, as with the storage controller, and those located in the CPU main memory. Experimental results are based on extensive trace-driven simulations using traces taken from three large IBM or IBM-compatible mainframe data processing installations. We find that disk cache is a powerful means of extending the performance limits of high-end computer systems.

An Adaptive Block Management Scheme Using On-Line Detection Of Block Reference Patterns Recent research has shown that near optimal performance can be achieved by adaptive block replacement policies that use user-level hints regarding the block reference pattern. However obtaining user-level hints requires considerable effort from users making it difficult to apply adaptive replacement policies to diverse kinds of applications. We propose a new adaptive black management scheme that we call DEAR (DEtection based Adaptive Replacement) which makes on-line detections of block reference patterns of applications using Decision Trees without user intervention. Based on the detected reference pattern, DEAR applies an appropriate replacement policy to each application. This scheme is suitable for buffer management in systems such as multimedia servers where data reference patterns of applications may be diverse. Results from trace driven simulations show that the DEAR scheme can detect the reference patterns of applications and reduce the miss ratio lip to 15 percentage points compared to the LRU policy.

Using dynamic sets to overcome high I/O latencies during search Describes a single unifying abstraction called 'dynamic sets', which can offer substantial benefits to search applications. These benefits include greater opportunity in the I/O subsystem to aggressively exploit prefetching and parallelism, as well as support for associative naming to complement the hierarchical naming in typical file systems. This paper motivates dynamic sets and presents the design of a system that embodies this abstraction.

Architectures and optimization methods of flash memory based storage systems Flash memory is a non-volatile memory which can be electrically erased and reprogrammed. Its major advantages such as small physical size, no mechanical components, low power consumption, and high performance have made it likely to replace the magnetic disk drives in more and more systems. However, flash memory has four specific features which are different to the magnetic disk drives, and pose challenges to develop practical techniques: (1) Flash memory is erased in blocks, but written in pages. (2) A block has to be erased before writing data to the block. (3) A block of flash memory can only be written for a specified number of times. (4) Writing pages within a block should be done sequentially. This survey presents the architectures, technologies, and optimization methods employed by the existing flash memory based storage systems to tackle the challenges. I hope that this paper will encourage researchers to analyze, optimize, and develop practical techniques to improve the performance and reduce the energy consumption of flash memory based storage systems, by leveraging the existing methods and solutions.

On Variable Scope of Parity Protection in Disk Arrays In a common form of a RAID 5 architecture, data is organized on a disk array consisting of N + 1 disks into stripes of N data blocks and one parity block (with parity block locations staggered so as to balance the number of parity blocks on each disk). This allows data to be recovered in the event of a single disk failure. Here we consider an extension to this architecture in which parity information applies to arbitrary subsets of the data blocks in each stripe. Using several simplifying assumptions, we present simulation and analytic results that provide estimates of the improvement using this approach, in terms of total I/O operations, as compared to 1) conventional RAID 5 under a random single-block write workload, and 2) the use of a log-structured file system in which data is written out in stripes. Results on the reduction of disk recovery costs are also presented.

Integration of buffer management and query optimization in relational database environment

Practical prefetching techniques for multiprocessor file systems Improvements in the processing speed of multiprocessors are outpacing improvements in the speed of disk hardware. Parallel disk I/O subsystems have been proposed as one way to close the gap between processor and disk speeds. In a previous paper we showed that prefetching and caching have thepotential to deliver the performance benefits of parallel file systems to parallel applications. In this paper we describe experiments withpractical prefetching policies that base decisions only on on-line reference history, and that can be implemented efficiently. We also test the ability of those policies across a range of architectural parameters.

Bridge: A High-Performance File System for Parallel Processors Faster storage devices cannot solve the I/O bottleneck problem forlargemultiprocessorsystemsifdatapassesthroughafllesys- temonasingleprocessor. Implementingthefllesystemasapar- allel program can signiflcantly improve performance. Selectively revealing this parallel structure to utility programs can produce additional improvements, particularly on machines in which in- terprocessor communication is slow compared to aggregate I/O bandwidth.

A cost-benefit scheme for high performance predictive prefetching

Notes on Data Base Operating Systems This paper is a compendium of data base management operating systems folklore. It is an early paper and is still in draft form. It is intended as a set of course notes for a class on data base operating systems. After a brief overview of what a data management system is it focuses on particular issues unique to the transaction management component especially locking and recovery.

Caching less for better performance: balancing cache size and update cost of flash memory cache in hybrid storage systems Hybrid storage solutions use NAND flash memory based Solid State Drives (SSDs) as non-volatile cache and traditional Hard Disk Drives (HDDs) as lower level storage. Unlike a typical cache, internally, the flash memory cache is divided into cache space and overprovisioned space, used for garbage collection. We show that balancing the two spaces appropriately helps improve the performance of hybrid storage systems. We show that contrary to expectations, the cache need not be filled with data to the fullest, but may be better served by reserving space for garbage collection. For this balancing act, we present a dynamic scheme that further divides the cache space into read and write caches and manages the three spaces according to the workload characteristics for optimal performance. Experimental results show that our dynamic scheme improves performance of hybrid storage solutions up to the off-line optimal performance of a fixed partitioning scheme. Furthermore, as our scheme makes efficient use of the flash memory cache, it reduces the number of erase operations thereby extending the lifetime of SSDs.

Open World Planning in the Situation Calculus We describe a forward reasoning planner for openworlds that uses domain specific information for pruningits search space, as suggested by (Bacchus & Kabanza1996; 2000). The planner is written in the situationcalculus-based programming language GOLOG,and it uses a situation calculus axiomatization of theapplication domain. Given a sentence oe to prove, theplanner regresses it to an equivalent sentence oe 0 aboutthe initial situation, then invokes a theorem prover todetermine...

DMP3: A Dynamic Multilayer Perceptron Construction Algorithm This paper presents a method for constructing multilayer perceptron networks (MLPs) called DMP3 (Dynamic Multilayer Perceptron 3). DMP3 differs from other MLP construction techniques in several important ways. The motivation for these differences and how they can lead to improved performance are discussed in detail in this paper. The DMP3 algorithm constructs MLPs by incrementally adding network elements to the output node of the network. Dependent upon the reduction in network error, the complexity of new elements that are added to the network can increase slightly with each growth cycle of the algorithm. As new elements are added to the network, the existing network structure is frozen and only the weights of the new elements are trained. In addition, the weights which link the new elements to the existing network structure are initially set to predetermined values, which predisposes each new network element to perform a particular function in relation to the existing network structure which can decrease the amount of time required for training the new elements. Information gain rather than error minimization is used to guide the growth of the network, which increases the utility of newly added network elements and decreases the likelihood that a premature dead end in the growth of the network will occur. A short, improvement driven training cycle is used to train new network elements which naturally helps to prevent over learning and memorization. The performance of DMP3 is compared with that of several other well-know machine learning and neural network learning algorithms (c4.5, cn2, ib1, CV based MLP architecture selection, c4, id3, perceptron, and mml) on 9 real world data sets taken from the UCI machine learning database. Simulation results show that DMP3 performs better (on average) than any of the other algorithms on the data sets tested.

Learning A Lexical Simplifier Using Wikipedia In this paper we introduce a new lexical simplification approach. We extract over 30K candidate lexical simplifications by identifying aligned words in a sentence-aligned corpus of English Wikipedia with Simple English Wikipedia. To apply these rules, we learn a feature-based ranker using SVMnk trained on a set of labeled simplifications collected using Amazon's Mechanical Turk. Using human simplifications for evaluation, we achieve a precision of 76% with changes in 86% of the examples.

Statistical and incremental methods for neural models selection This work presents two methods of selection of neural models for identification of dynamic systems. Initially, a strategy of selection based on statistical tests, which relates to training and generalisation performances of a neural model is analysed. In the second time, a new constructive approach of neural model selection, which the training begins with minimal structure and then incrementally adds new hidden units and/or layers, is described. The simulation and the application of these methods for selection of neural models are also considered.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

A study of I/O system organizations With the increasing processing speeds, it has become important to design powerful and efficient I/O systems. In this paper, we look at several design options in designing an I/O system and study their impact on the performance. Specifically, we use trace driven simulations to study a disk system with a nonvolatile cache. Some of the considered design parameters include the cache block size, the fetch size, the cache size and the disk access policy. We show that decoupling the fetch size and the cache block size results in significant performance improvements. A new write-back policy is presented that is shown to offer significant performance benefits. We show that optimal block size in a two-level memory hierarchy is dependent only on the latency, data rate product of the second level as previously conjectured. We also present results showing the effect of a split access operation of a disk read/write head.

Providing QoS guarantees for disk I/O In this paper, we address the problem of providing different levels of performance guarantees or quality ofservice for disk I/O. We classify disk requests into threecategories based on the provided level of service. We propose an integrated scheme that provides different levels ofperformance guarantees in a single system. We propose andevaluate a mechanism for providing deterministic servicefor variable-bit-rate streams at the disk. We will show that,through proper admission control and bandwidth allocation,requests in different categories can be ensured of performance guarantees without getting impacted by requests inother categories. We evaluate the impact of scheduling policy decisions on the provided service. We also quantify theimprovements in stream throughput possible by using statistical guarantees instead of deterministic guarantees in thecontext of the proposed approach.

Operating system support for multimedia systems Distributed multimedia applications will be an important part of tomorrow's application mix and require appropriate operating system (OS) support. Neither hard real-time solutions nor best-effort solutions are directly well suited for this support. One reason is the co-existence of real-time and best effort requirements in future systems. Another reason is that the requirements of multimedia applications are not easily predictable, like variable bit rate coded video data and user interactivity. In this article, we present a survey of new developments in OS support for (distributed) multimedia systems, which include: (1) development of new CPU and disk scheduling mechanisms that combine real-time and best effort in integrated solutions; (2) provision of mechanisms to dynamically adapt resource reservations to current needs; (3) establishment of new system abstractions for resource ownership to account more accurate resource consumption; (4) development of new file system structures; (5) introduction of memory management mechanisms that utilize knowledge about application behavior; (6) reduction of major performance bottlenecks, like copy operations in I/O subsystems; and (7) user-level control of resources including communication.

Distributed schedule management in the Tiger video fileserver Tiger is a scalable, fault-tolerant video file server constructed from a collection of computers connected by a switclied network., All content files are striped across all of the computers and disks in a Tiger system. In order to prevent conflicts for a particular resource between two viewers, Tiger schedules viewers so that ihey do not require access to the same resource at the same time. In the abstract, there is a single, global schedule that describes all of the viewers in the system. In practice, the schedule is distrib&d among all of the computers in the system, each of which has a possibly partially inconsistent view of a subset of the schedule. By using such a relaxed consistency model for the schedule, Tiger achieves scalability and fault tolerance while still providing $e consistent, coordinated service required by viewers.

Failure evaluation of disk array organizations The authors present an evaluation of some of the disk array organizations proposed in the literature. They evaluate three alternatives for sparing, hot sparing, distributed sparing, and parity sparing, and two options for data layout, regular RAID5 and block designs, and systems based on combinations of these data layout and sparing alternatives. The performance of these organizations is evaluated with different reconstruction strategies. It is shown that parity sparing and distributed sparing have better performance and shorter reconstruction times than hot sparing. It is shown that both block designs as a data layout policy and distributed sparing as a sparing policy reduce the reconstruction time after a failure. The impact of reconstruction strategies is studied, and it is shown that, at higher workloads, choice of reconstruction strategy has a significant impact on the performance of the systems

Evolving mach 3.0 to a migrating thread model We have modified Mach 3.0 to treat cross-domain remote procedure call (RPC) as a single entity, instead of a sequence of message passing operations. With RPC thus elevated, we improved the transfer of control during RPC by changing the thread model. Like most operating systems, Mach views threads as statically associated with a single task, with two threads involved in an RPC. An alternate model is that of migrating threads, in which, during RPC, a single thread abstraction moves between tasks with the logical flow of control, and "server" code is passively executed. We have compatibly replaced Mach's static threads with migrating threads, in an attempt to isolate this aspect of operating system design and implementation. The key element of our design is a decoupling of the thread abstraction into the execution context and the schedulable thread of control, consisting of a chain of contexts. A key element of our implementation is that threads are now "based" in the kernel, and temporarily make excursions into tasks via upcalls. The new system provides more precisely defined semantics for thread manipulation and additional control operations, allows scheduling and accounting attributes to follow threads, simplifies kernel code, and improves RPC performance. We have retained the old thread and IPC interfaces for backwards compatibility, with no changes required to existing client programs and only a minimal change to servers, as demonstrated by a functional Unix single server and clients. The logical complexity along the critical RPC path has been reduced by a factor of nine. Local RPC, doing normal marshaling, has sped up by factors of 1.7-3.4. We conclude that a migrating-thread model is superior to a static model, that kernel-visible RPC is a prerequisite for this improvement, and that it is feasible to improve existing operating systems in this manner.

The Tiger Shark file system Tiger Shark is a parallel file system for IBM's AIX operating system. It is designed to support interactive multimedia, particularly large-scale systems such as interactive television (ITV). Tiger Shark scales across the entire RS/6000 product line, from small desktop machines to the SP-2 parallel supercomputer. Tiger Shark's primary features are support for continuous time data, scalability, high availability, and manageability, all of which are crucial in its role in large-scale video servers. Interestingly, most of the features that make Tiger Shark a good video server are important for other large-scale applications such as technical computing, data mining, digital library, and scalable network file servers. This paper briefly describes Tiger Shark: the environment that makes it important, the key technology it embodies, and the efforts to build products based on it.

Failure correction techniques for large disk arrays The ever increasing need for I/O bandwidth will be met with ever larger arrays of disks. These arrays require redundancy to protect against data loss. This paper examines alternative choices for encodings, or codes, that reliably store information in disk arrays. Codes are selected to maximize mean time to data loss or minimize disks containing redundant data, but are all constrained to minimize performance penalties associated with updating information or recovering from catastrophic disk failures. We also codes that give highly reliable data storage with low redundant data overhead for arrays of 1000 information disks.

Extended ephemeral logging: log storage management for applications with long lived transactions Extended ephemeral logging (XEL) is a new technique for managing a log of database activity subject to the general assumption that the lifetimes of an application’s transactions may be statistically distributed over a wide range. The log resides on nonvolatile disk storage and provides fault tolerance to system failures (in which the contents of volatile main memory storage may be lost). XEL segments a log into a chain of fixed-size FIFO queues and performs generational garbage collection on records in the log. Log records that are no longer necessary for recovery purposes are “thrown away” when they reach the head of a queue; only records that are still needed for recovery are forwarded from the head of one queue to the tail of the next. XEL does not require checkpoints, permits fast recovery after a crash and is well suited for applications that have a wide distribution of transaction lifetimes. Quantitative evaluation of XEL via simulation indicates that it can significantly reduce the disk space required for the log, at the expense of slightly higher bandwidth for log information and more main memory; the reduced size of the log permits much faster recovery after a crash as well as cost savings. XEL can significantly reduce both the disk space and the disk bandwidth required for log information in a system that has been augmented with a nonvolatile region of main memory.

Fault tolerant design of multimedia servers Recent technological advances have made multimedia on-demand servers feasible. Two challenging tasks in such systems are: a) satisfying the real-time requirement for continuous delivery of objects at specified bandwidths and b) efficiently servicing multiple clients simultaneously. To accomplish these tasks and realize economies of scale associated with servicing a large user population, the multimedia server can require a large disk subsystem. Although a single disk is fairly reliable, a large disk farm can have an unacceptably high probability of disk failure. Further, due to the real-time constraint, the reliability and availability requirements of multimedia systems are very stringent. In this paper we investigate techniques for providing a high degree of reliability and availability, at low disk storage, bandwidth, and memory costs for on-demand multimedia servers.

SOPA: Selecting the optimal caching policy adaptively With the development of storage technology and applications, new caching policies are continuously being introduced. It becomes increasingly important for storage systems to be able to select the matched caching policy dynamically under varying workloads. This article proposes SOPA, a cache framework to adaptively select the matched policy and perform policy switches in storage systems. SOPA encapsulates the functions of a caching policy into a module, and enables online policy switching by policy reconstruction. SOPA then selects the policy matched with the workload dynamically by collecting and analyzing access traces. To reduce the decision-making cost, SOPA proposes an asynchronous decision making process. The simulation experiments show that no single caching policy performed well under all of the different workloads. With SOPA, a storage system could select the appropriate policy for different workloads. The real-system evaluation results show that SOPA reduced the average response time by up to 20.3% and 11.9% compared with LRU and ARC, respectively.

Input space versus feature space in kernel-based methods. This paper collects some ideas targeted at advancing our understanding of the feature spaces associated with support vector (SV) kernel functions. We first discuss the geometry of feature space. In particular, we review what is known about the shape of the image of input space under the feature space map, and how this influences the capacity of SV methods. Following this, we describe how the metric governing the intrinsic geometry of the mapped surface can be computed in terms of the kernel, using the example of the class of inhomogeneous polynomial kernels, which are often used in SV pattern recognition. We then discuss the connection between feature space and input space by dealing with the question of how one can, given some vector in feature space, find a preimage (exact or approximate) in input space. We describe algorithms to tackle this issue, and show their utility in two applications of kernel methods. First, we use it to reduce the computational complexity of SV decision functions; second, we combine it with the Kernel PCA algorithm, thereby constructing a nonlinear statistical denoising technique which is shown to perform well on real-world data.

Utilizing Problem Structure in Planning: A Local Search Approach

Learning Topic Representation For Smt With Neural Networks Statistical Machine Translation (SMT) usually utilizes contextual information to disambiguate translation candidates. However, it is often limited to contexts within sentence boundaries, hence broader topical information cannot be leveraged. In this paper, we propose a novel approach to learning topic representation for parallel data using a neural network architecture, where abundant topical contexts are embedded via topic relevant monolingual data. By associating each translation rule with the topic representation, topic relevant rules are selected according to the distributional similarity with the source text during SMT decoding. Experimental results show that our method significantly improves translation accuracy in the NIST Chinese-to-English translation task compared to a state-of-the-art baseline.

ParoC++: A Requirement-driven Parallel Object-oriented Programming Language Adaptive utilization of resources in a highly heterogeneous computational environment such as the Grid is a difficult question. In this paper we address an object-oriented approach to the solution using requirement-driven parallel objects. Each parallel object is a self-described, shareable and passive object that resides in a separate memory address space. The allocation of the parallel object is driven by the constraints on the resource on which the object will live. A new parallel programming paradigm is presented in the context of ParoC++ - a new parallel object-oriented programming environment for high performance distributed computing. ParoC++ extends C++ for supporting requirement-driven parallel objects and a runtime system that provides services to run ParoC+ + programs in distributed environments. An industrial application on realtime image processing is used as a test case to the system. The experimental results show that the ParoC++ model is efficient and scalable and that it makes easier to adapt parallel applications to dynamic environments.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Cross-modal Retrieval with Correspondence Autoencoder The problem of cross-modal retrieval, e.g., using a text query to search for images and vice-versa, is considered in this paper. A novel model involving correspondence autoencoder (Corr-AE) is proposed here for solving this problem. The model is constructed by correlating hidden representations of two uni-modal autoencoders. A novel optimal objective, which minimizes a linear combination of representation learning errors for each modality and correlation learning error between hidden representations of two modalities, is used to train the model as a whole. Minimization of correlation learning error forces the model to learn hidden representations with only common information in different modalities, while minimization of representation learning error makes hidden representations are good enough to reconstruct input of each modality. A parameter $\\alpha$ is used to balance the representation learning error and the correlation learning error. Based on two different multi-modal autoencoders, Corr-AE is extended to other two correspondence models, here we called Corr-Cross-AE and Corr-Full-AE. The proposed models are evaluated on three publicly available data sets from real scenes. We demonstrate that the three correspondence autoencoders perform significantly better than three canonical correlation analysis based models and two popular multi-modal deep models on cross-modal retrieval tasks.

A data-driven study of image feature extraction and fusion Feature analysis is the extraction and comparison of signals from multimedia data, which can subsequently be semantically analyzed. Feature analysis is the foundation of many multimedia computing tasks such as object recognition, image annotation, and multimedia information retrieval. In recent decades, considerable work has been devoted to the research of feature analysis. In this work, we use large-scale datasets to conduct a comparative study of four state-of-the-art, representative feature extraction algorithms: color-texture codebook (CT), SIFT codebook, HMAX, and convolutional networks (ConvNet). Our comparative evaluation demonstrates that different feature extraction algorithms enjoy their own advantages, and excel in different image categories. We provide key observations to explain where these algorithms excel and why. Based on these observations, we recommend feature extraction principles and identify several pitfalls for researchers and practitioners to avoid. Furthermore, we determine that in a large training dataset with more than 10,000 instances per image category, the four evaluated algorithms can converge to the same high level of category-prediction accuracy. This result supports the effectiveness of the data-driven approach. Finally, based on learned clues from each algorithm's confusion matrix, we devise a fusion algorithm to harvest synergies between these four algorithms and further improve class-prediction accuracy.

Learning semantic representation with neural networks for community question answering retrieval •Learning the semantic representation using neural network architecture.•The neural network is trained via pre-training and fine-tuning phase.•The learned semantic level feature is incorporated into a LTR framework.

Learning Compact Face Representation: Packing a Face into an int32 This paper addresses the problem of producing very compact representation of a face image for large-scale face search and analysis tasks. In tradition, the compactness of face representation is achieved by a dimension reduction step after representation extraction. However, the dimension reduction usually degrades the discriminative ability of the original representation drastically. In this paper, we present a deep learning framework which optimizes the compactness and discriminative ability jointly. The learnt representation can be as compact as 32 bit (same as the int32) and still produce highly discriminative performance (91.4% on LFW benchmark). Based on the extreme compactness, we show that traditional face analysis tasks (e.g. gender analysis) can be effectively solved by a Look-Up-Table approach given a large-scale face data set.

3D Mesh Labeling via Deep Convolutional Neural Networks This article presents a novel approach for 3D mesh labeling by using deep Convolutional Neural Networks (CNNs). Many previous methods on 3D mesh labeling achieve impressive performances by using predefined geometric features. However, the generalization abilities of such low-level features, which are heuristically designed to process specific meshes, are often insufficient to handle all types of meshes. To address this problem, we propose to learn a robust mesh representation that can adapt to various 3D meshes by using CNNs. In our approach, CNNs are first trained in a supervised manner by using a large pool of classical geometric features. In the training process, these low-level features are nonlinearly combined and hierarchically compressed to generate a compact and effective representation for each triangle on the mesh. Based on the trained CNNs and the mesh representations, a label vector is initialized for each triangle to indicate its probabilities of belonging to various object parts. Eventually, a graph-based mesh-labeling algorithm is adopted to optimize the labels of triangles by considering the label consistencies. Experimental results on several public benchmarks show that the proposed approach is robust for various 3D meshes, and outperforms state-of-the-art approaches as well as classic learning algorithms in recognizing mesh labels.

Multimodal Deep Autoencoder for Human Pose Recovery Video-based human pose recovery is usually conducted by retrieving relevant poses using image features. In the retrieving process, the mapping between 2D images and 3D poses is assumed to be linear in most of the traditional methods. However, their relationships are inherently non-linear, which limits recovery performance of these methods. In this paper, we propose a novel pose recovery method usi...

Efficient Learning of Domain-invariant Image Representations

Extreme Learning Classifier with Deep Concepts.

Intentions and attention in exploratory health search We study information goals and patterns of attention in explorato-ry search for health information on the Web, reporting results of a large-scale log-based study. We examine search activity associated with the goal of diagnosing illness from symptoms versus more general information-seeking about health and illness. We decom-pose exploratory health search into evidence-based and hypothe-sis-directed information seeking. Evidence-based search centers on the pursuit of details and relevance of signs and symptoms. Hypothesis-directed search includes the pursuit of content on one or more illnesses, including risk factors, treatments, and therapies for illnesses, and on the discrimination among different diseases under the uncertainty that exists in advance of a confirmed diag-nosis. These different goals of exploratory health search are not independent, and transitions can occur between them within or across search sessions. We construct a classifier that identifies medically-related search sessions in log data. Given a set of search sessions flagged as health-related, we show how we can identify different intentions persisting as foci of attention within those sessions. Finally, we discuss how insights about foci dynamics can help us better understand exploratory health search behavior and better support health search on the Web.

Big Data Deep Learning: Challenges and Perspectives Deep learning is currently an extremely active research area in machine learning and pattern recognition society. It has gained huge successes in a broad area of applications such as speech recognition, computer vision, and natural language processing. With the sheer size of data available today, big data brings big opportunities and transformative potential for various sectors; on the other hand, it also presents unprecedented challenges to harnessing data and information. As the data keeps getting bigger, deep learning is coming to play a key role in providing big data predictive analytics solutions. In this paper, we provide a brief overview of deep learning, and highlight current research efforts and the challenges to big data, as well as the future trends.

Automated planning

Probabilistic data exchange The work reported here lays the foundations of data exchange in the presence of probabilistic data. This requires rethinking the very basic concepts of traditional data exchange, such as solution, universal solution, and the certain answers of target queries. We develop a framework for data exchange over probabilistic databases, and make a case for its coherence and robustness. This framework applies to arbitrary schema mappings, and finite or countably infinite probability spaces on the source and target instances. After establishing this framework and formulating the key concepts, we study the application of the framework to a concrete and practical setting where probabilistic databases are compactly encoded by means of annotations formulated over random Boolean variables. In this setting, we study the problems of testing for the existence of solutions and universal solutions, materializing such solutions, and evaluating target queries (for unions of conjunctive queries) in both the exact sense and the approximate sense. For each of the problems, we carry out a complexity analysis based on properties of the annotation, in various classes of dependencies. Finally, we show that the framework and results easily and completely generalize to allow not only the data, but also the schema mapping itself to be probabilistic.

A Framework for Adaptive Storage Input/Output on Computational Grids Emerging computational grids consist of distributed collections of heterogeneous sequential and parallel systems and irregular applications with complex, data dependent execution behavior and time varying resource demands. To provide adaptive input/output resource management for these systems, we are developing PPFS II, a portable parallel file system. PPFS II supports rule-based, closed loop and interactive control of input/output subsystems on both parallel and wide area distributed systems.

Red-black planning: A new systematic approach to partial delete relaxation. To date, delete relaxation underlies some of the most effective heuristics for deterministic planning. Despite its success, however, delete relaxation has significant pitfalls in many important classes of planning domains, and it has been a challenge from the outset to devise heuristics that take some deletes into account. We herein devise an elegant and simple method for doing just that. In the context of finite-domain state variables, we define red variables to take the relaxed semantics, in which they accumulate their values rather than switching between them, as opposed to black variables that take the regular semantics. Red–black planning then interpolates between relaxed planning and regular planning simply by allowing a subset of variables to be painted red. We investigate the tractability region of red–black planning, extending Chen and Giménez' characterization theorems for regular planning to the more general red–black setting. In particular, we identify significant islands of tractable red–black planning, use them to design practical heuristic functions, and experiment with a range of “painting strategies” for automatically choosing the red variables. Our experiments show that these new heuristic functions can improve significantly on the state of the art in satisficing planning.1

Interpolating Destin Features For Image Classification This paper presents a novel approach for image classification, by integrating advanced machine learning techniques and the concept of feature interpolation. In particular, a recently introduced learning architecture, the Deep Spatio-Temporal Inference Network (DeSTIN) [1], is employed to perform feature extraction for support vector machine (SVM) based image classification. The system is supported by use of a simple interpolation mechanism, which allows the improvement of the original low-dimensionality of feature sets to a significantly higher dimensionality with minimal computation. This in turn, improves the performance of SVM classifiers while reducing the computation otherwise required to generate directly measured features. The work is tested against the popular MNIST dataset of handwritten digits [2]. Experimental results indicate that the proposed approach is highly promising, with the integrated system generally outperforming that which makes use of pure DeSTIN as the feature extraction preprocessor to SVM classifiers.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Analysis of Agent Programs Using Action Models The use of action models for the analysis of control programs can be useful for two reasons. First, it promises to deliver better tools for the simulation, verification and synthesis of control programs, and second it presents challenging problems for theories of action and knowledge. In this paper we use a theory of actions and knowledge developed elsewhere to analyze control programs for navigation tasks.We model both physical and sensing actions and establish conditions under which different control programs are executable and lead the agent to the intended goal.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

A simple declarative language for describing narratives with actions We describe a simple declarative languageEfor describing the effects of a series of action occurrences within a narrative.Eis analogous to Gelfond and Lifschitz's LanguageAand its extensions, but is based on a different ontology. The semantics ofEis based on a simple characterisation of persistence which facilitates a modular approach to extending the expressivity of the language. Domain descriptions inAcan be translated to equivalent theories inE. We show how, in the context of reasoning about actions,E's narrative-based ontology may be exploited in order to characterise and synthesise two complementary notions of explanation. According to the first notion, explanation may be partly modelled as the process of suitably extending an apparently inconsistent theory written inEso as to establish consistency, thus providing a natural method, in many cases, to account for conflicting sets of information about the domain. According to the second notion, observations made at later times can sometimes be explained in terms of what is true at earlier times. This enables domains to be given an alternative characterisation in which knowledge arising from observations is appropriately separated from other aspects of the domain. We also describe howEdomains may be implemented as Event Calculus style logic programs, which facilitate automated reasoning both backwards and forwards in time, and which behave correctly even when the knowledge entailed by the domain description is incomplete.

A Realistic Experiment in Knowledge Representation in Open Event Calculus: Protocol Specification This paper presents one of the first realistic experiments in the use of Event Calculusin Open Logic Programming: the specification of a process protocol. The specification taskinvolves most of the common complications of temporal reasoning: the representation of contextdependent actions, of preconditions and ramifications of actions, the modelling of systemfaults, and most of all, the representation of uncertainty of actions. As the underlying language,the Open Logic Programming...

E-RES: A System for Reasoning about Actions, Events and Observations E-RES is a system that implements the Language E, a logic for reasoning about narratives of action oc- currences and observations. E's semantics is model- theoretic, but this implementation is based on a sound and complete reformulation of E in terms of argumen- tation, and uses general computational techniques of argumentation frameworks. The system derives scep- tical non-monotonic consequences of a given reformu- lated theory which exactly correspond to consequences entailed by E's model-theory. The computation relies on a complimentary ability of the system to derive cred- ulous non-monotonic consequences together with a set of supporting assumptions which is sucient for the (credulous) conclusion to hold. E-RES allows theories to contain general action laws, statements about ac- tion occurrences, observations and statements of ram- ications (or universal laws). It is able to derive con- sequences both forward and backward in time. This paper gives a short overview of the theoretical basis of E-RES and illustrates its use on a variety of examples. Currently, E-RES is being extended so that the system can be used for planning.

Reasoning about Actions, Narratives and Ramification The Language E is a simple declarative language for describingthe effects of action occurrences within a given narrative, usingan ontology of actions, time points and fluents (i.e. propertieswhich can change their truth values over time). This paper showshow E may be extended to deal with ramifications. More precisely,we show how Language E domain descriptions can includestatements describing permanent relationships or constraints betweenfluents, and how the model theoretic semantics of...

Causality and the Qualification Problem In formal theories for reasoning about actions, the qualification problem denotes the problem to account for the many conditions which, albeit being unlikely to occur, may prevent the successful execution of an action. By a simple counter-example in the spirit of the well-known Yale Shooting scenario, we show that the common straightforward approach of globally minimizing such abnormal disqualifications is inadequate as it lacks an appropriate notion of causality. To overcome this difficulty, we propose to incorporate causality by treating the proposition that an action is qualified as a fluent which is initially assumed away by default but otherwise potentially indirectly affected by the execution of actions. Our formal account of the qualification problem includes the proliferation of explanations for surprising disqualifications and also accommodates so-called miraculous disqualifications. We moreover sketch a version of the fluent calculus which involves default rules to address abnormal disqualifications of actions, and which is provably correct wrt. our formal characterization of the qualification problem.

Logic Programming and Negation: A Survey. We survey here various approaches which were proposed to incorporate negation in logicprograms. We concentrate on the proof-theoretic and model-theoretic issues and the relationshipsbetween them.1991 Mathematics Subject Classification: 68Q40, 68T15.CR Categories: F.3.2., F.4.1, H.3.3, I.2.3.Keywords and Phrases: negation, general logic programs, non-monotonic reasoning.Notes. The work of the first author was partly supported by ESPRIT Basic Research Action6810 (Compulog 2). The work...

A logic-based calculus of events Formal Logic can be used to represent knowledge of many kinds for many purposes. It can be used to formalize programs, program specifications, databases, legislation, and natural language in general. For many such applications of logic a representation of time is necessary. Although there have been several attempts to formalize the notion of time in classical first-order logic, it is still widely believed that classical logic is not adequate for the representation of time and that some form of non-classical Temporal Logic is needed. In this paper, we shall outline a treatment of time, based on the notion of event, formalized in the Horn clause subset of classical logic augmented with negation as failure. The resulting formalization is executable as a logic program. We use the term ''event calculus'' to relate it to the well-known ''situation calculus'' (McCarthy and Hayes 1969). The main difference between the two is conceptual: the situation calculus deals with global states whereas the event calculus deals with local events and time periods. Like the event calculus, the situation calculus can be formalized by means of Horn clauses augmented with negation by failure (Kowalski 1979). The main intended applications investigated in this paper are the updating of data- bases and narrative understanding. In order to treat both cases uniformly we have taken the view that an update consists of the addition of new knowledge to a knowledge base. The effect of explicit deletion of information in conventional databases is obtained without deletion by adding new knowledge about the end of the period of time for which the information holds.

A logic programming approach to knowledge-state planning, II: the DLVk system In Part I of this series of papers, we have proposed a new logic-based planning language, called K. This language facilitates the description of transitions between states of knowledge and it is well suited for planning under incomplete knowledge. Nonetheless,K also supports the representation of transitions between states of the world (i.e., states of complete knowledge) as a special case, proving to be very flexible. In the present Part II, we describe the DLVK planning system, which implements K on top of the disjunctive logic programming system DLV. This novel planning system allows for solving hard planning problems, including secure planning under incomplete initial states (often called conformant planning in the literature), which cannot be solved at all by other logic-based planning systems such as traditional satisfiability planners. We present a detailed comparison of the DLVK system to several state-of-the-art conformant planning systems, both at the level of system features and on benchmark problems. Our results indicate that, thanks to the power of knowledge-state problem encoding, the DLVK system is competitive even with special purpose conformant planning systems, and it often supplies a more natural and simple representation of the planning problems.

From Causal Theories to Successor State Axioms and STRIPS-Like Systems We describe a system for specifying the efiects of actions. Unlike those commonly used in AI planning, our system uses an action description language that allows one to specify the efiects of actions using domain rules, which are state con- straints that can entail new action efiects from old ones. Declaratively, an action domain in our lan- guage corresponds to a nonmonotonic causal the- ory in the situation calculus. Procedurally, such an action domain is compiled into a set of proposi- tional theories, one for each action in the domain, from which fully instantiated successor state-like axioms and STRIPS-like systems are then gener- ated. We expect the system to be a useful tool for knowledge engineers writing action speciflca- tions for classical AI planning systems, GOLOG systems, and other systems where formal specifl- cations of actions are needed.

Conformant Planning via Model Checking . Conformant planning is the problem of nding a sequenceof actions that is guaranteed to achieve the goal for any possible initialstate and nondeterministic behavior of the planning domain. In this paperwe present a new approach to conformant planning. We propose analgorithm that returns the set of all conformant plans of minimal lengthif the problem admits a solution, otherwise it returns with failure. Ourwork is based on the planning via model checking paradigm, and relieson...

Some connections between bounded query classes and non-uniform complexity It is shown that if there is a polynomial-time algorithm that tests k(n)=O(log n) points for membership in a set A by making only k(n)-1 adaptive queries to an oracle set X, then A belongs to NP/poly intersection co-NP/poly (if k(n)=O(1) then A belong to P/poly). In particular, k(n)=O(log n) queries to an NP -complete set (k(n)=O(1) queries to an NP-hard set) are more powerful than k(n)-1 queries, unless the polynomial hierarchy collapses. Similarly, if there is a small circuit that tests k(n) points for membership in A by making only k(n)-1 adaptive queries to a set X, then there is a correspondingly small circuit that decides membership in A without an oracle. An investigation is conducted of the quantitatively stronger assumption that there is a polynomial-time algorithm that tests 2k strings for membership in A by making only k queries to an oracle X, and qualitatively stronger conclusions about the structure of A are derived: A cannot be self-reducible unless A∈P, and A cannot be NP-hard unless P=NP. Similar results hold for counting classes. In addition, relationships between bounded-query computations, lowness, and the p-degrees are investigated

Learning a class of large finite state machines with a recurrent neural network One of the issues in any learning model is how it scales with problem size. The problem of learning finite state machine (FSMs) from examples with recurrent neural networks has been extensively explored. However, these results are somewhat disappointing in the sense that the machines that can be learned are too small to be competitive with existing grammatical inference algorithms. We show that a type of recurrent neural network (Narendra & Parthasarathy, 1990, IEEE Trans. Neural Networks, 1 , 4–27) which has feedback but no hidden state neurons can learn a special type of FSM called a finite memory machine (FMM) under certain constraints. These machines have a large number of states (simulations are for 256 and 512 state FMMs) but have minimal order, relatively small depth and little logic when the FMM is implemented as a sequential machine.

Read Optimized File System Designs: A Performance Evaluation This paper presents a performance comparison of several file system allocation policies. The file systems are designed to provide high bandwidth between disks and main memory by taking advantage of parallelism in an underlying disk array, catering to large units of transfer, and minimizing the bandwidth dedicated to the transfer of meta data. All of the file systems described use a mul- tiblock allocation strategy which allows both large and small files to be allocated efficiently. Simulation results show that these multiblock policies result in systems that are able to utilize a large percentage of the underlying disk bandwidth; more than 90% in sequential cases. As general purpose systems are called upon to support more data intensive applications such as databases and super- computing, these policies offer an opportunity to provide superior performance to a larger class of users.

Improving Citation Polarity Classification With Product Reviews Recent work classifying citations in scientific literature has shown that it is possible to improve classification results with extensive feature engineering. While this result confirms that citation classification is feasible, there are two drawbacks to this approach: (i) it requires a large annotated corpus for supervised classification, which in the case of scientific literature is quite expensive; and (ii) feature engineering that is too specific to one area of scientific literature may not be portable to other domains, even within scientific literature. In this paper we address these two drawbacks. First, we frame citation classification as a domain adaptation task and leverage the abundant labeled data available in other domains. Then, to avoid over-engineering specific citation features for a particular scientific domain, we explore a deep learning neural network approach that has shown to generalize well across domains using unigram and bigram features. We achieve better citation classification results with this cross-domain approach than using in-domain classification.

Fine mapping of key soil nutrient content using high resolution remote sensing image to support precision agriculture in Northwest China The rapid development of industrialized agriculture has leads to the problems of soil pollution and water pollution. In order to solve these problems, precision agriculture (PA) has been applied to achieve precise management of agricultural water and fertilizer. In PA process, fine mapping of soil nutrient is an effective technology to acquire accurate water and fertilizer distribution information and make agricultural decision. A significant progress has been made in digital soil mapping (DSM) of soil nutrient content over the past 20 years. However, the accuracy of grid-based DSM cannot meet the practical application needs of PA. This paper proposed a fine DSM method of soil nutrient content using high resolution remote sensing images and multi-scale auxiliary data for PA application. Three key technologies were studied for the implementation of this method. The automatic extraction of fine mapping units was the basis of this method. We designed different automatic extraction methods based on high resolution remote sensing images for agricultural production units in plains and mountainous areas. The auxiliary variables in different scales were chosen and converted to construct fine-scale soil nutrient-environment relationship model. Finally, machine learning methods were used to map the spatial distribution of soil nutrients. We chose Zhongning County, Ningxia Province as the study area, which includes typical plain and mountainous agriculture. The proposed method and technologies were applied for typical soil nutrients mapping. A common grid-based spatial interpolation method was implemented with the same soil sample dataset to evaluate the effect of the proposed method. The result showed that this method could reduce the number of prediction units and effectively improve the prediction efficiency in both plain and mountainous areas for fine soil mapping and precision agriculture application. This study was an attempt to realize fine soil mapping based on PA application unit in different environments. The high-resolution remote sensing images provide basic data for the realization of this idea, and the conversion technology of multi-scale data provides better support for the spatial inference of fine soil attribute information. In the future, we will carry out experiments in larger areas to further improve the efficiency of application, and plan to expand this study to consider three-dimensional soil property prediction.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Learning Topic Representation For Smt With Neural Networks Statistical Machine Translation (SMT) usually utilizes contextual information to disambiguate translation candidates. However, it is often limited to contexts within sentence boundaries, hence broader topical information cannot be leveraged. In this paper, we propose a novel approach to learning topic representation for parallel data using a neural network architecture, where abundant topical contexts are embedded via topic relevant monolingual data. By associating each translation rule with the topic representation, topic relevant rules are selected according to the distributional similarity with the source text during SMT decoding. Experimental results show that our method significantly improves translation accuracy in the NIST Chinese-to-English translation task compared to a state-of-the-art baseline.

A Survey: Time Travel in Deep Learning Space: An Introduction to Deep Learning Models and How Deep Learning Models Evolved from the Initial Ideas This report will show the history of deep learning evolves. It will trace back as far as the initial belief of connectionism modelling of brain, and come back to look at its early stage realization: neural networks. With the background of neural network, we will gradually introduce how convolutional neural network, as a representative of deep discriminative models, is developed from neural networks, together with many practical techniques that can help in optimization of neural networks. On the other hand, we will also trace back to see the evolution history of deep generative models, to see how researchers balance the representation power and computation complexity to reach Restricted Boltzmann Machine and eventually reach Deep Belief Nets. Further, we will also look into the development history of modelling time series data with neural networks. We start with Time Delay Neural Networks and move further to currently famous model named Recurrent Neural Network and its extension Long Short Term Memory. We will also briefly look into how to construct deep recurrent neural networks. Finally, we will conclude this report with some interesting open-ended questions of deep neural networks.

Why are deep nets reversible: A simple theory, with implications for training Generative models for deep learning are promising both to improve understanding of the model, and yield training methods requiring fewer labeled samples. Recent works use generative model approaches to produce the deep net's input given the value of a hidden layer several levels above. However, there is no accompanying "proof of correctness" for the generative model, showing that the feedforward deep net is the correct inference method for recovering the hidden layer given the input. Furthermore, these models are complicated. The current paper takes a more theoretical tack. It presents a very simple generative model for RELU deep nets, with the following characteristics: (i) The generative model is just the reverse of the feedforward net: if the forward transformation at a layer is $A$ then the reverse transformation is $A^T$. (This can be seen as an explanation of the old weight tying idea for denoising autoencoders.) (ii) Its correctness can be proven under a clean theoretical assumption: the edge weights in real-life deep nets behave like random numbers. Under this assumption ---which is experimentally tested on real-life nets like AlexNet--- it is formally proved that feed forward net is a correct inference method for recovering the hidden layer. The generative model suggests a simple modification for training: use the generative model to produce synthetic data with labels and include it in the training set. Experiments are shown to support this theory of random-like deep nets; and that it helps the training.

Provable Bounds for Learning Some Deep Representations. We give algorithms with provable guarantees that learn a class of deep nets in the generative model view popularized by Hinton and others. Our generative model is an $n$ node multilayer neural net that has degree at most $n^{\gamma}$ for some $\gamma <1$ and each edge has a random edge weight in $[-1,1]$. Our algorithm learns {\em almost all} networks in this class with polynomial running time. The sample complexity is quadratic or cubic depending upon the details of the model. The algorithm uses layerwise learning. It is based upon a novel idea of observing correlations among features and using these to infer the underlying edge structure via a global graph recovery procedure. The analysis of the algorithm reveals interesting structure of neural networks with random edge weights.

Train faster, generalize better: Stability of stochastic gradient descent We show that parametric models trained by a stochastic gradient method (SGM) with few iterations have vanishing generalization error. We prove our results by arguing that SGM is algorithmically stable in the sense of Bousquet and Elisseeff. Our analysis only employs elementary tools from convex and continuous optimization. We derive stability bounds for both convex and non-convex optimization under standard Lipschitz and smoothness assumptions. Applying our results to the convex case, we provide new insights for why multiple epochs of stochastic gradient methods generalize well in practice. In the non-convex case, we give a new interpretation of common practices in neural networks, and formally show that popular techniques for training large deep models are indeed stability-promoting. Our findings conceptually underscore the importance of reducing training time beyond its obvious benefit.

Why Does Unsupervised Pre-training Help Deep Learning? Much recent research has been devoted to learning algorithms for deep architectures such as Deep Belief Networks and stacks of auto-encoder variants, with impressive results obtained in several areas, mostly on vision and language data sets. The best results obtained on supervised learning tasks involve an unsupervised learning component, usually in an unsupervised pre-training phase. Even though these new algorithms have enabled training deep models, many questions remain as to the nature of this difficult learning problem. The main quest ion investigated here is the following: how does unsupervised pre-training work? Answering this questions is important if learning in deep architectures is to be further improved. We propose several explanatory hypotheses and test them through extensive simulations. We empirically show the influence of pre-training with respect to architecture depth, model capacity, and number of training examples. The experiments confirm and clarify the advantage of unsupervised pre-training. The results suggest that unsupervised pre- training guides the learning towards basins of attraction o f minima that support better generalization from the training data set; the evidence from these results s upports a regularization explanation for the effect of pre-training.

Logic Programming and Negation: A Survey. We survey here various approaches which were proposed to incorporate negation in logicprograms. We concentrate on the proof-theoretic and model-theoretic issues and the relationshipsbetween them.1991 Mathematics Subject Classification: 68Q40, 68T15.CR Categories: F.3.2., F.4.1, H.3.3, I.2.3.Keywords and Phrases: negation, general logic programs, non-monotonic reasoning.Notes. The work of the first author was partly supported by ESPRIT Basic Research Action6810 (Compulog 2). The work...

On the Desirability of Acyclic Database Schemes A class of database schemes, called acychc, was recently introduced. It is shown that this class has a number of desirable properties. In particular, several desirable properties that have been studied by other researchers m very different terms are all shown to be eqmvalent to acydicity. In addition, several equivalent charactenzauons of the class m terms of graphs and hypergraphs are given, and a smaple algorithm for determining acychclty is presented. Also given are several eqmvalent characterizations of those sets M of multivalued dependencies such that M is the set of muRlvalued dependencies that are the consequences of a given join dependency. Several characterizations for a conflict-free (in the sense of Lien) set of muluvalued dependencies are provided.

Consensus and Cooperation in Networked Multi-Agent Systems? This paper provides a theoretical framework for analysis of consensus algorithms for multi-agent networked systems with an emphasis on the role of directed information flow, robustness to changes in network topology due to link/node failures, time-delays, and performance guarantees. An overview of basic concepts of information consensus in networks and methods of convergence and performance analys...

Reasoning about action I: a possible worlds approach Reasoning about change is an important aspect of commonsense reasoning and planning.In this paper we describe an approach to reasoning about change for rich domains whereit is not possible to anticipate all situations that might occur. The approach provides asolution to the frame problem, and to the related problem that it is not always reasonable toexplicitly specify all of the consequences of actions. The approach involves keeping a singlemodel of the world that is updated when actions...

Serverless network file systems We propose a new paradigm for network file system design: serverless network file systems. While traditional network file systems rely on a central server machine, a serverless system utilizes workstations cooperating as peers to provide all file system services. Any machine in the system can store, cache, or control any block of data. Our approach uses this location independence, in combination with fast local area networks, to provide better performance and scalability than traditional file systems. Furthermore, because any machine in the system can assume the responsibilities of a failed component, our serverless design also provides high availability via redundatn data storage. To demonstrate our approach, we have implemented a prototype serverless network file system called xFS. Preliminary performance measurements suggest that our architecture achieves its goal of scalability. For instance, in a 32-node xFS system with 32 active clients, each client receives nearly as much read or write throughput as it would see if it were the only active client.

Scheduling a mixed interactive and batch workload on a parallel, shared memory supercomputer

Reasoning About Actions in Narrative Understanding Reasoning about actions has been a focus of interest in AI from the beginning and continues to receive attention. Rut the range of situations considered has been rather narrow and falls well short of what is needed for understanding natural language. Language understanding requires sophisticated reasoning about actions and events and the world's languages employ a variety of grammatical and lexical devices to construe, direct attention and focus on, and control inferences about actions and events. We implemented a neurally inspired computational model that is able to reason about, linguistic action and event descriptions, such as those found in news stories. The system uses an active. event representation that also seems to provide natural and cognitiveIy motivated solutions to classical problems in logical theories of reasoning about actions. For logical approaches to reasoning about actions, we suggest, that looking at story understanding sets up fairly strong desiderata both in terms of the fine-grained event and action distinctions and the kinds of real-time inferences required.

Wsben: A Web Services Discovery And Composition Benchmark Toolkit In this article, a novel benchmark toolkit, WSBen, for testing web services discovery and composition algorithms is presented. The WSBen includes: (1) a collection of synthetically generated web services files in WSDL format with diverse data and model characteristics; (2) queries for testing discovery and composition algorithms; (3) auxiliary files to do statistical analysis on the WSDL test sets; (4) converted WSDL test sets that conventional AI planners can read; and (5) a graphical interface to control all these behaviors. Users can fine-tune the generated WSDL test files by varying underlying network models. To illustrate the application of the WSBen, in addition, we present case studies from three domains: (1) web service composition; (2) AI planning; and (3) the laws of networks in Physics community. It is our hope that WSBen will provide useful insights in evaluating the performance of web services discovery and composition algorithms. The WSBen toolkit is available at: http://pike.psu.edu/sw/wsben/.

Learning A Lexical Simplifier Using Wikipedia In this paper we introduce a new lexical simplification approach. We extract over 30K candidate lexical simplifications by identifying aligned words in a sentence-aligned corpus of English Wikipedia with Simple English Wikipedia. To apply these rules, we learn a feature-based ranker using SVMnk trained on a set of labeled simplifications collected using Amazon's Mechanical Turk. Using human simplifications for evaluation, we achieve a precision of 76% with changes in 86% of the examples.

Mining Minimal Non-redundant Association Rules Using Frequent Closed Itemsets The problem of the relevance and the usefulness of extracted association rules is of primary importance because, in the majority of cases, real-life databases lead to several thousands association rules with high confidence and among which are many redundancies. Using the closure of the Galois connection, we define two new bases for association rules which union is a generating set for all valid association rules with support and confidence. These bases are characterized using frequent closed itemsets and their generators; they consist of the nonredundant exact and approximate association rules having minimal antecedents and maximal consequents, i.e. the most relevant association rules. Algorithms for extracting these bases are presented and results of experiments carried out on real-life databases show that the proposed bases are useful, and that their generation is not time consuming.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

Concurrent actions in the situation calculus We propose a representation of Concurrent actions; rather than invent a new formalism, we model them within the standard situation calculus by introducing the notions of global actions and primitive actions, whose relationship is analogous to' that between situations and fluents. The result is a framework in which situations and actions play quite symmetric roles. The rich structure of actions gives rise to' a new problem, which, due to' this symmetry between actions and situations, is analogous to' the traditional frame problem. In [Lin and Shoham 1991] we provided a solution to' the frame problem based on a formal adequacy criterion called "epistemological completeness." Here we show how to' solve the new problem based on the same adequacy criterion.

Exploring Gate-Limited Analytical Models for High Performance Network Storage Servers

Parameterized complexity for the database theorist

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

HFS: a performance-oriented flexible file system based on building-block compositions The Hurricane File System (HFS) is designed for (potentially large-scale) shared-memory multiprocessors. Its architecture is based on the principle that, in order to maximize performance for applications with diverse requirements, a file system must support a wide variety of file structures, file system policies, and I/O interfaces. Files in HFS are implemented using simple building blocks composed in potentially complex ways. This approach yields great flexibility, allowing an application to customize the structure and policies of a file to exactly meet its requirements. As an extreme example, HFS allows a file's structure to be optimized for concurrent random-access write-only operations by 10 threads, something no other file system can do. Similarly, the prefetching, locking, and file cache management policies can all be chosen to match an application's access pattern. In contrast, most parallel file systems support a single file structure and a small set of policies. We have implemented HFS as part of the Hurricane operating system running on the Hector shared-memory multiprocessor. We demonstrate that the flexibility of HFS comes with little processing or I/O overhead. We also show that for a number of file access patterns, HFS is able to deliver to the applications the full I/O bandwidth of the disks on our system.

Flexibility and performance of parallel file systems As we gain experience with parallel file systems, it becomes increasingly clear that a single solution does not suit all applications. For example, it appears to be impossible to find a single appropriate interface, caching policy, file structure, or disk-management strategy. Furthermore, the proliferation of file-system interfaces and abstractions make applica- tions difficult to port. We propose that the traditional functionality of parallel file systems be separated into two components: a fixed core that is standard on all plat- forms, encapsulating only primitive abstractions and interfaces, and a set of high-level libraries to provide a variety of abstractions and application- programmer interfaces (APIs). We present our current and next-generation file systems as examples of this structure. Their features, such as a three-dimensional file structure, strided read and write interfaces, and I/O-node programs, are specifically designed with the flexibility and performance necessary to support a wide range of applications.

Dynamic I/O characterization of I/O intensive scientific applications Understanding the characteristic I/O behavior of scientific applications is an integral part of the research and development efforts for the improvement of high performance I/O systems. This study focuses on application level I/O behavior with respect to both static and dynamic characteristics. We observed the San Diego Supercomputer Center's Cray C90 workload and isolated the most I/O intensive applications. The combination of a low-level description of physical resource usage and the high-level functional composition of applications and scientific disciplines for this set reveals the major sources of I/O demand in the workload. We selected two applications from the I/O intensive set and performed a detailed analysis of their dynamic I/O behavior. These applications exhibited a high degree of regularity in their I/O activity over time and their characteristic I/O behaviors can be precisely described by one and two, respectively, recurring sequences of data accesses and computation periods. Key Words: empirical I/O behavior, supercomputer applications

Tuning the performance of I/O-intensive parallel applications Getting good I/O performance from parallel programs is a critical problem for many application domains. Inthis paper, we report our experience tuning the I/O performance of four application programs from the areas ofsensor data processing and linear algebra. After tuning, three of the four applications achieve effective I/O rates ofover 100Mb/s, on 16 processors. The total volume of I/O required by the programs ranged from about 75MB toover 200GB. We report the lessons learned in achieving...

PPFS: a high performance portable parallel file system Rapid increases in processor performance over the pastdecade have outstripped performance improvements ininput/output devices, increasing the importance of input/output performance to overall system performance.Further, experience has shown that the performance ofparallel input/output systems is particularly sensitive todata placement and data management policies, makinggood choices critical. To explore this vast design space, wehave developed a user-level library, the Portable...

Tolerating latency through software-controlled prefetching in shared-memory multiprocessors The large latency of memory accesses is a major obstacle in obtaining high processor utilization in large scale shared-memory multiprocessors. Although the provision of coherent caches in many recent machines has alleviated the problem somewhat, cache misses still occur frequently enough that they significantly lower performance. In this paper we evaluate the effectiveness of non-binding software-controlled prefetching , as proposed in the Stanford DASH Multiprocessor, to address this problem. The prefetches are non-binding in the sense that the prefetched data is brought to a cache close to the processor, but is still available to the cache coherence protocol to keep it consistent. Prefetching is software-controlled since the program must explicitly issue prefetch instructions. The paper presents results from detailed simulation studies done in the context of the Stanford DASH multiprocessor. Our results show that for applications with regular data access patterns—we evaluate a particle- based simulator used in aeronautics and an LU-decomposition application—prefetching can be very effective. It was easy to augment the applications to do prefetching and it increased their performance by 100-150% when we prefetched directly into the processor's cache. However, for applications with complex data usage patterns, prefetching was less successful. After much effort, the performance of a distributed-time logic simulation application that made extensive use of pointers and linked lists could be increased only by 30%. The paper also evaluates the effects of various hardware optimizations such as separate prefetch issue buffers, prefetching with exclusive ownership, lockup-free caches, and weaker memory consistency models on the performance of prefetching.

A study of integrated prefetching and caching strategies Prefetching and caching are effective techniques for improving the performance of file systems, but they have not been studied in an integrated fashion. This paper proposes four properties that optimal integrated strategies for prefetching and caching must satisfy, and then presents and studies two such integrated strategies, called aggressive and conservative. We prove that the performance of the conservative approach is within a factor of two of optimal and that the performance of the aggressive strategy is a factor significantly less than twice that of the optimal case. We have evaluated these two approaches by trace-driven simulation with a collection of file access traces. Our results show that the two integrated prefetching and caching strategies are indeed close to optimal and that these strategies can reduce the running time of applications by up to 50%.

Intelligent caching for remote file service Limitations of current disk block caching strategies are discussed. A model for providing remote file service using knowledge-based caching algorithms is proposed. The knowledge-based algorithms generate expectations of user process behavior which are used to provide hints to the file server. Te research involved gathering trace data from a modified Unix kernel and conducting trace-driven simulation of remote file server models. Performance improvements of up to 340% were observed for knowledge-based caching in simulated file service. Comparisons are made between conventional, knowledge-based, and optimal models. Extensions to general caching are discussed

Multiple-Level MPI File Write-Back and Prefetching for Blue Gene Systems This paper presents the design and implementation of an asynchronous data-staging strategy for file accesses based on ROMIO, the most popular MPI-IO distribution, and ZeptoOS, an open source operating system solution for Blue Gene systems. We describe and evaluate a two-level file write-back implementation and a one-level prefetching solution. The experimental results demonstrate that both solutions achieve high performance through a high degree of overlap between computation, communication, and file I/O.

Background data movement in a log-structured disk subsystem The log-structured disk subsystem is a new concept for the use of disk storage whose future application has enormous potential. In such a subsystem, all writes are organized into a log, each entry of which is placed into the next available free storage. A directory indicates the physical location of each logical object (e.g., each file block or track image) as known to the processor originating the I/O request. For those objects that have been written more than once, the directory retains the location of the most recent copy. Other work with log-structured disk subsystems has shown that they are capable of high write throughputs. However, the fragmentation of free storage due to the scattered locations of data that become out of date can become a problem in sustained operation. To control fragmentation, it is necessary to perform ongoing garbage collection, in which the location of stored data is shifted to release unused storage for re-use. This paper introduces a mathematical model of garbage collection, and shows how collection load relates to the utilization of storage and the amount of locality present in the pattern of updates. A realistic statistical model of updates, based upon trace data analysis, is applied. In addition, alternative policies are examined for determining which data areas to collect. The key conclusion of our analysis is that in environments with the scattered update patterns typical of database I/O, the utilization of storage must be controlled in order to achieve the high write throughput of which the subsystem is capable. In addition, the presence of data locality makes it important to take the past history of data into account in determining the next area of storage to be garbage-collected.

RISC: A resilient interconnection network for scalable cluster storage systems The explosive growth of data generated by information digitization has been identified as the key driver to escalate storage requirements. It is becoming a big challenge to design a resilient and scalable interconnection network which consolidates hundreds even thousands of storage nodes to satisfy both the bandwidth and storage capacity requirements. This paper proposes a resilient interconnection network for storage cluster systems (RISC). The RISC divides storage nodes into multiple partitions to facilitate the data access locality. Multiple spare links between any two storage nodes are employed to offer strong resilience to reduce the impact of the failures of links, switches, and storage nodes. The scalability is guaranteed by plugging in additional switches and storage nodes without reconfiguring the overall system. Another salient feature is that the RISC achieves a dynamic scalability of resilience by expanding the partition size incrementally with additional storage nodes along with associated two network interfaces that expand resilience degree and balance workload proportionally. A metric named resilience coefficient is proposed to measure the interconnection network. A mathematical model and the corresponding case study are employed to illustrate the practicability and efficiency of the RISC.

Optimization Problems And The Polynomial Hierarchy It is demonstrated that such problems as the symmetric Travelling Salesman Problem, Chromatic Number Problem, Maximal Clique Problem and a Knapsack Packing Problem are in the Δ P 2 level of PH and no lower if ∑ P 1 ≠ Π P 1 , or NP≠co-NP. This shows that these problems cannot be solved by polynomial reductions that use only positive information from an NP oracle, if NP≠co-NP. It is then shown how to extend these results to prove that interesting problems are properly in Δ P, X i +1 for all X , k where ∑ P, X k ≠ Π P, X k in PH X .

Planning with Different Forms of Domain-Dependent Control Knowledge - An Answer Set Programming Approach In this paper we present a declarative approach to adding domain-dependent control knowledge for Answer Set Planning (ASP). Our approach allows different types of domain-dependent control knowledge such as hierarchical, temporal, or procedural knowledge to be represented and exploited in parallel, thus combining the ideas of control knowledge in HTN-planning, GOLOG-programming, and planning with temporal knowledge into ASP. To do so, we view domain-dependent control knowledge as sets of independent constraints. An advantage of this approach is that domain-dependent control knowledge can be modularly formalized and added to the planning problem as desired. We define a set of constructs for constraint representation and provide a set of domain-independent logic programming rules for checking constraint satisfaction.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Single pass streaming BLAST on FPGAs Approximate string matching is fundamental to bioinformatics and has been the subject of numerous FPGA acceleration studies. We address issues with respect to FPGA implementations of both BLAST- and dynamic-programming- (DP) based methods. Our primary contribution is a new algorithm for emulating the seeding and extension phases of BLAST. This operates in a single pass through a database at streaming rate, and with no preprocessing other than loading the query string. Moreover, it emulates parameters turned to maximum possible sensitivity with no slowdown. While current DP-based methods also operate at streaming rate, generating results can be cumbersome. We address this with a new structure for data extraction. We present results from several implementations showing order of magnitude acceleration over serial reference code. A simple extension assures compatibility with NCBI BLAST.

Speeding up subset seed algorithm for intensive protein sequence comparison Abstract—Sequence similarity search is a common and re- peated task in molecular biology. The rapid growth of genomic databases leads to the need of speeding up the treatment of this task. In this paper, we present a subset seed algorithm for intensive protein sequence comparison. We have accelerated this algorithm by using indexing technique and fine grained parallelism of GPU and SIMD instructions. We have implemented two programs: iBLASTP, iTBLASTN. The GPU (SIMD) imple- mentation of the two programs achieves a speed up ranging from 5.5 to 10 (4 to 5.6) compared to the BLASTP and TBLASTN of the BLAST program family, with comparable sensitivity.

The Astral Compendium For Protein Structure And Sequence Analysis The ASTRAL compendium provides several databases and tools to aid in the analysis of protein structures, particularly through the use of their sequences. The SPACI scores included in the system summarize the overall characteristics of a protein structure. A structural alignments database indicates residue equivalencies in superimposed protein domain structures, The PDB sequence-map files provide a linkage between the amino acid sequence of the molecule studied (SEQRES records in a database entry) and the sequence of the atoms experimentally observed in the structure (ATOM records). These maps are combined with information in the SCOP database to provide sequences of protein domains. Selected subsets of the domain database, with varying degrees of similarity measured in several different ways, are also available. ASTRAL may be accessed at http://astral.stanford.edu/.

Accelerating BLASTP on the Cell Broadband Engine The enormous growth of biological sequence databases has caused bioinformatics to be rapidly moving towards a data-intensive, computational science. As a result, the computational power needed by bioinformatics applications is growing rapidly as well. The recent emergence of low cost parallel accelerator technologies has made it possible to reduce execution times of many bioinformatics applications. In this paper, we demonstrate how the PlayStation®3, powered by the Cell Broadband Engine, can be used as an efficient computational platform to accelerate the popular BLASTP algorithm.

Proceedings of the 24th International Conference on Supercomputing, 2010, Tsukuba, Ibaraki, Japan, June 2-4, 2010

A rate-based prefiltering approach to blast acceleration DNA sequence comparison and database search have evolved in the last years as a field of strong competition between several reconfigurable hardware computing groups. In this paper we present a BLAST preprocessor that efficiently marks the parts of the database that may produce matches. Our prefiltering approach offers significant reduction in the size of the database that needs to be fully processed by BLAST, with a corresponding reduction in the run-time of the algorithm. We have implemented our architecture, evaluated its effectiveness for a variety of databases and queries, and compared its accuracy against the original NCBI Blast implementation. We have found that prefiltering offers at least a factor of 5 and up to 3 orders of magnitude reduction in the database space that needs to be fully searched. Due to its prefiltering nature, our approach can be combined with all major reconfigurable acceleration architectures that have been presented up to date.

High-throughput sequence alignment using Graphics Processing Units. Background: The recent availability of new, less expensive high-throughput DNA sequencing technologies has yielded a dramatic increase in the volume of sequence data that must be analyzed. These data are being generated for several purposes, including genotyping, genome resequencing, metagenomics, and de novo genome assembly projects. Sequence alignment programs such as MUMmer have proven essential for analysis of these data, but researchers will need ever faster, high-throughput alignment tools running on inexpensive hardware to keep up with new sequence technologies. Results: This paper describes MUMmerGPU, an open-source high-throughput parallel pairwise local sequence alignment program that runs on commodity Graphics Processing Units (GPUs) in common workstations. MUMmerGPU uses the new Compute Unified Device Architecture (CUDA) from nVidia to align multiple query sequences against a single reference sequence stored as a suffix tree. By processing the queries in parallel on the highly parallel graphics card, MUMmerGPU achieves more than a 10-fold speedup over a serial CPU version of the sequence alignment kernel, and outperforms the exact alignment component of MUMmer on a high end CPU by 3.5-fold in total application time when aligning reads from recent sequencing projects using Solexa/Illumina, 454, and Sanger sequencing technologies. Conclusion: MUMmerGPU is a low cost, ultra-fast sequence alignment program designed to handle the increasing volume of data produced by new, high-throughput sequencing technologies. MUMmerGPU demonstrates that even memory-intensive applications can run significantly faster on the relatively low-cost GPU than on the CPU.

The Swiss-Prot Protein Knowledgebase And Its Supplement Trembl In 2003 The SWISS-PROT protein knowledgebase (http: / / www. expasy. org/ sprot/ and http: / / www. ebi. ac. uk/ swissprot/) connects amino acid sequences with the current knowledge in the Life Sciences. Each protein entry provides an interdisciplinary overview of relevant information by bringing together experimental results, computed features and sometimes even contradictory conclusions. Detailed expertise that goes beyond the scope of SWISS-PROT is made available via direct links to specialised databases. SWISS-PROT provides annotated entries for all species, but concentrates on the annotation of entries from human ( the HPI project) and other model organisms to ensure the presence of high quality annotation for representative members of all protein families. Part of the annotation can be transferred to other family members, as is already done for microbes by the High-quality Automated and Manual Annotation of microbial Proteomes (HAMAP) project. Protein families and groups of proteins are regularly reviewed to keep up with current scientific findings. Complementarily, TrEMBL strives to comprise all protein sequences that are not yet represented in SWISS-PROT, by incorporating a perpetually increasing level of mostly automated annotation. Researchers are welcome to contribute their knowledge to the scientific community by submitting relevant findings to SWISS-PROT at [email protected].

Predictive rule inference for epistatic interaction detection in genome-wide association studies. Under the current era of genome-wide association study (GWAS), finding epistatic interactions in the large volume of SNP data is a challenging and unsolved issue. Few of previous studies could handle genome-wide data due to the difficulties in searching the combinatorially explosive search space and statistically evaluating high-order epistatic interactions given the limited number of samples. In this work, we propose a novel learning approach (SNPRuler) based on the predictive rule inference to find disease-associated epistatic interactions.Our extensive experiments on both simulated data and real genome-wide data from Wellcome Trust Case Control Consortium (WTCCC) show that SNPRuler significantly outperforms its recent competitor. To our knowledge, SNPRuler is the first method that guarantees to find the epistatic interactions without exhaustive search. Our results indicate that finding epistatic interactions in GWAS is computationally attainable in practice.http://bioinformatics.ust.hk/SNPRuler.zip

A study of replacement algorithms for a virtual-storage computer One of the basic limitations of a digitalcomputer is the size of its available memory.'I n most cases, it is neither feasible nor economical for a user to insist that every problem program fit into memory. The number of words of information in a pro-gramoften exceeds the number of cells (i. e., word locations) in memory. The only way to solve this problem is to assign more than one program word to a cell. Since a cell can hold only one word at a time, extra words assigned tothe cell must be held inexternalstorage. Conventionally, overlay techniques are em-ployed to exchange memory words and external-storage words whenever needed; this, of course, places an additional planning and coding burden on the programmer. For several reasons, it wouldbe advantageous to rid the programmer of thisfunction by providing him witha'virtual" memory larger than his pro-gram. An approach that permits him to use a sufficiently large address range can accomplish this objective, assuming that means are provided for automatic execution of the memory-overlay functions. Among the first and most promising of the large-address approaches is the one described byKilburn, et a1.'A similar for framework, the relative merits of various specific algorithms are compared. Before

Distributed operating systems Distributed operating systems have many aspects in common with centralized ones, but they also differ in certain ways. This paper is intended as an introduction to distributed operating systems, and especially to current university research about them. After a discussion of what constitutes a distributed operating system and how it is distinguished from a computer network, various key design issues are discussed. Then several examples of current research projects are examined in some detail, namely, the Cambridge Distributed Computing System, Amoeba, V, and Eden.

Nonparametric belief propagation for self-localization of sensor networks Automatic self-localization is a critical need for the effective use of ad hoc sensor networks in military or civilian applications. In general, self-localization involves the combination of absolute location information (e.g., from a global positioning system) with relative calibration information (e.g., distance measurements between sensors) over regions of the network. Furthermore, it is generally desirable to distribute the computational burden across the network and minimize the amount of intersensor communication. We demonstrate that the information used for sensor localization is fundamentally local with regard to the network topology and use this observation to reformulate the problem within a graphical model framework. We then present and demonstrate the utility of nonparametric belief propagation (NBP), a recent generalization of particle filtering, for both estimating sensor locations and representing location uncertainties. NBP has the advantage that it is easily implemented in a distributed fashion, admits a wide variety of statistical models, and can represent multimodal uncertainty. Using simulations of small to moderately sized sensor networks, we show that NBP may be made robust to outlier measurement errors by a simple model augmentation, and that judicious message construction can result in better estimates. Furthermore, we provide an analysis of NBP's communications requirements, showing that typically only a few messages per sensor are required, and that even low bit-rate approximations of these messages can be used with little or no performance impact.

Raising a Hardness Result This article presents a technique for proving problems hard for classes of the polynomial hierarchy or for PSPACE. The rationale of this technique is that some problem restrictions are able to simulate existential or universal quantifiers. If this is the case, reductions from Quantified Boolean Formulae (QBF) to these restrictions can be transformed into reductions from QBFs having one more quantifier in the front. This means that a proof of hardness of a problem at level n in the polynomial hierarchy can be split into n separate proofs, which may be simpler than a proof directly showing a reduction from a class of QBFs to the considered problem.

Learning Topic Representation For Smt With Neural Networks Statistical Machine Translation (SMT) usually utilizes contextual information to disambiguate translation candidates. However, it is often limited to contexts within sentence boundaries, hence broader topical information cannot be leveraged. In this paper, we propose a novel approach to learning topic representation for parallel data using a neural network architecture, where abundant topical contexts are embedded via topic relevant monolingual data. By associating each translation rule with the topic representation, topic relevant rules are selected according to the distributional similarity with the source text during SMT decoding. Experimental results show that our method significantly improves translation accuracy in the NIST Chinese-to-English translation task compared to a state-of-the-art baseline.

Ensemble learning with trees and rules: Supervised, semi-supervised, unsupervised In this article, we propose several new approaches for post processing a large ensemble of conjunctive rules for supervised, semi-supervised and unsupervised learning problems. We show with various examples that for high dimensional regression problems the models constructed by post processing the rules with partial least squares regression have significantly better prediction performance than the ones produced by the random forest or the rulefit algorithms which use equal weights or weights estimated from lasso regression. When rule ensembles are used for semi-supervised and unsupervised learning, the internal and external measures of cluster validity point to high quality groupings.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Reasoning about actions with sensing under qualitative and probabilistic uncertainty We focus on the aspect of sensing in reasoning about actions under qualitative and probabilistic uncertainty. We first define the action language E for reasoning about actions with sensing, which has a semantics based on the autoepistemic description logic ALCKNF, and which is given a formal semantics via a system of deterministic transitions between epistemic states. As an important feature, the main computational tasks in E can be done in linear and quadratic time. We then introduce the action language E+ for reasoning about actions with sensing under qualitative and probabilistic uncertainty, which is an extension of E by actions with nondeterministic and probabilistic effects, and which is given a formal semantics in a system of deterministic, nondeterministic, and probabilistic transitions between epistemic states. We also define the notion of a belief graph, which represents the belief state of an agent after a sequence of deterministic, nondeterministic, and probabilistic actions, and which compactly represents a set of unnormalized probability distributions. Using belief graphs, we then introduce the notion of a conditional plan and its goodness for reasoning about actions under qualitative and probabilistic uncertainty. We formulate the problems of optimal and threshold conditional planning under qualitative and probabilistic uncertainty, and show that they are both uncomputable in general. We then give two algorithms for conditional planning in our framework. The first one is always sound, and it is also complete for the special case in which the relevant transitions between epistemic states are cycle-free. The second algorithm is a sound and complete solution to the problem of finite-horizon conditional planning in our framework. Under suitable assumptions, it computes every optimal finite-horizon conditional plan in polynomial time. We also describe an application of our formalism in a robotic-soccer scenario, which underlines its usefulness in realistic applications.

Probabilistic Situation Calculus In this article we propose a Probabilistic Situation Calculus logical language to represent and reason with knowledge about dynamic worlds in which actions have uncertain effects. Uncertain effects are modeled by dividing an action into two subparts: a deterministic (agent produced) input and a probabilistic reaction (produced by nature). We assume that the probabilities of the reactions have known distributions.Our logical language is an extension to Situation Calculae in the style proposed by Raymond Reiter. There are three aspects to this work. First, we extend the language in order to accommodate the necessary distinctions (e.g., the separation of actions into inputs and reactions). Second, we develop the notion of Randomly Reactive Automata in order to specify the semantics of our Probabilistic Situation Calculus. Finally, we develop a reasoning system in MATHEMATICA capable of performing temporal projection in the Probabilistic Situation Calculus.

A Logical Framework to Reinforcement Learning Using Hybrid Probabilistic Logic Programs Knowledge representation is an important issue in reinforcement learning. Although logic programming with answer set semantics is a standard in knowledge representation, it has not been exploited in reinforcement learning to resolve its knowledge representation issues. In this paper, we present a logic programming framework to reinforcement learning, by integrating reinforcement learning, in MDP environments, with normal hybrid probabilistic logic programs with probabilistic answer set semantics [29], that is capable of representing domain-specific knowledge. We show that any reinforcement learning problem, MT, can be translated into a normal hybrid probabilistic logic program whose probabilistic answer sets correspond to trajectories in MT. We formally prove the correctness of our approach. Moreover, we show that the complexity of finding a policy for a reinforcement learning problem in our approach is NP-complete. In addition, we show that any reinforcement learning problem, MT, can be encoded as a classical logic program with answer set semantics, whose answer sets corresponds to valid trajectories in MT. We also show that a reinforcement learning problem can be encoded as a SAT problem. In addition, we present a new high level action description language that allows the factored representation of MDP.

Reasoning about actions with imprecise and incomplete state descriptions This article is a first step in the direction of extending possibilistic planning to account for incomplete and imprecise knowledge of the world state. Fundamental definitions are given and the possibilistic planning problem is recast in this new setting. Finally, it is shown that, under certain conditions, possibilistic planning with imprecise and incomplete state descriptions is no harder than possibilistic planning with crisp and complete information.

Logic, Knowledge Representation, and Bayesian Decision Theory In this paper I give a brief overview of recent work on uncertainty in AI, and relate it to logical representations. Bayesian decision theory and logic are both normative frameworks for reasoning that emphasize different aspects of intelligent reasoning. Belief networks (Bayesian networks) are representations of independence that form the basis for understanding much of the recent work on reasoning under uncertainty, evidential and causal reasoning, decision analysis, dynamical systems, optimal control, reinforcement learning and Bayesian learning. The independent choice logic provides a bridge between logical representations and belief networks that lets us understand these other representations and their relationship to logic and shows how they can extended to first-order rule-based representations. This paper discusses what the representations of uncertainty can bring to the computational logic community and what the computational logic community can bring to those studying reasoning under uncertainty.

A Semantical Account of Progression in the Presence of Defaults In previous work, we proposed a modal fragment of the situation calculus called ${\mathcal ES}$, which fully captures Reiter's basic action theories. ${\mathcal ES}$ also has epistemic features, including only-knowing, which refers to all that an agent knows in the sense of having a knowledge base. While our model of only-knowing has appealing properties in the static case, it appears to be problematic when actions come into play. First of all, its utility seems to be restricted to an agent's initial knowledge base. Second, while it has been shown that only-knowing correctly captures default inferences, this was only in the static case, and undesirable properties appear to arise in the presence of actions. In this paper, we remedy both of these shortcomings and propose a new dynamic semantics of only-knowing, which is closely related to Lin and Reiter's notion of progression when actions are performed and where defaults behave properly.

Complexity of finite-horizon Markov decision process problems Controlled stochastic systems occur in science engineering, manufacturing, social sciences, and many other cntexts. If the systems is modeled as a Markov decision process (MDP) and will run ad infinitum, the optimal control policy can be computed in polynomial time using linear programming. The problems considered here assume that the time that the process will run is finite, and based on the size of the input. There are mny factors that compound the complexity of computing the optimal policy. For instance, there are many factors that compound the complexity of this computation. For instance, if the controller does not have complete information about the state of the system, or if the system is represented in some very succint manner, the optimal policy is provably not computable in time polynomial in the size of the input. We analyze the computational complexity of evaluating policies and of determining whether a sufficiently good policy exists for a MDP, based on a number of confounding factors, including the observability of the system state; the succinctness of the representation; the type of policy; even the number of actions relative to the number of states. In almost every case, we show that the decision problem is complete for some known complexity class. Some of these results are familiar from work by Papadimitriou and Tsitsiklis and others, but some, such as our PL-completeness proofs, are surprising. We include proofs of completeness for natural problems in the as yet little-studied classes NPPP.

The complexity of stochastic games We consider the complexity of stochastic games—simple games of chance played by two players. We show that the problem of deciding which player has the greatest chance of winning the game is in the class NP ⌢ co- NP .

Planning with Incomplete Information as Heuristic Search in Belief Space The formulation of planning as heuristic search withheuristics derived from problem representations hasturned out to be a fruitful approach for classical planning.In this paper, we pursue a similar idea in thecontext planning with incomplete information. Planningwith incomplete information can be formulated asa problem of search in belief space, where belief statescan be either sets of states or more generally probabilitydistribution over states. While the formulation (as the...

Parallel non-binary planning in polynomial time This paper formally presents a class of planning problems which allows non-binary state variables and parallel execution of actions. The class is proven to be tractable, and we provide a sound and complete polynomial time algorithm for planning within this class. This result means that we are getting closed to tackling realistic planning problems in sequential control, where a restricted problem representation is often sufficient, but where the size of the problems make tractability an important issue.

Measurements of a distributed file system We analyzed the user-level file access patterns and caching behavior of the Sprite distributed file system. The first part of our analysis repeated a study done in 1985 of the: BSD UNIX file system. We found that file throughput has increased by a factor of 20 to an average of 8 Kbytes per second per active user over 10-minute intervals, and that the use of process migration for load sharing increased burst rates by another factor of six. Also, many more very large (multi-megabyte) files are in use today than in 1985. The second part of our analysis measured the behavior of Sprite's main-memory file caches. Client-level caches average about 7 Mbytes in size (about one-quarter to one-third of main memory) and filter out about 50% of the traffic between clients and servers. 35% of the remaining server traffic is caused by paging, even on workstations with large memories. We found that client cache consistency is needed to prevent stale data errors, but that it is not invoked often enough to degrade overall system performance.

Architectural implications of quantum computing technologies In this article we present a classification scheme for quantum computing technologies that is based on the characteristics most relevant to computer systems architecture. The engineering trade-offs of execution speed, decoherence of the quantum states, and size of systems are described. Concurrency, storage capacity, and interconnection network topology influence algorithmic efficiency, while quantum error correction and necessary quantum state measurement are the ultimate drivers of logical clock speed. We discuss several proposed technologies. Finally, we use our taxonomy to explore architectural implications for common arithmetic circuits, examine the implementation of quantum error correction, and discuss cluster-state quantum computation.

Enhancing write I/O performance of disk array RM2 tolerating double disk failures With a large number of internal disks and the rapid growth of disk capacity, storage systems become more susceptible to double disk failures. Thus, the need for such reliable storage systems as RAID6 is expected to gain in importance. However RAID6 architectures such as RM2, P+Q, EVEN-ODD, and DATUM traditionally suffer from a low write I/O performance caused by updating two distinctive parity data associated with user data. To overcome such a low write I/O performance, we propose an enhanced RM2 architecture which combines RM2, one of the well-known RAID6 architectures, with a Lazy Parity Update (LPU) technique. Extensive performance evaluations reveal that the write I/O performance of the proposed architecture is about two times higher than that of RM2 under various I/O workloads with little degradation in reliability.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

The LOCKSS peer-to-peer digital preservation system The LOCKSS project has developed and deployed in a world-wide test a peer-to-peer system for preserving access to journals and other archival information published on the Web. It consists of a large number of independent, low-cost, persistent Web caches that cooperate to detect and repair damage to their content by voting in “opinion polls.” Based on this experience, we present a design for and simulations of a novel protocol for voting in systems of this kind. It incorporates rate limitation and intrusion detection to ensure that even some very powerful adversaries attacking over many years have only a small probability of causing irrecoverable damage before being detected.

Keeping Bits Safe: How Hard Can It Be? As storage systems grow larger and larger, protecting their data for long-term storage is becoming more and more challenging.

Maximizing Throughput in Replicated Disk Striping of Variable Bit-Rate Streams In a system, offering on-demand real-time streaming of media files, data striping across an array of disks can improve load balancing, allowing higher disk utilization and increased system throughput. However, it can also cause complete service disruption in the case of a disk failure. Reliability can be improved by adding data redundancy and reserving extra disk bandwidth during normal operation. In this paper,, we are interested in providing fault-tolerance for media servers that support variable bit-rate encoding formats. Higher compression efficiency with respect to constant bit-rate encoding can significantly reduce per-user resource requirements, at the cost of increased resource management complexity. For the first time, the interaction between storage system fault-tolerance and variable bit-rate streaming with deterministic QoS guarantees is investigated. We implement into a prototype server and experimentally evaluate, using detailed simulated disk models', alternative data replication techniques and disk band, width reservation schemes. We show that with the minimum, reservation scheme introduced here, single disk failures can be tolerated at a cost Of less than 20% reduced throughput during normal operation, even for a disk array of moderate size. We also examine the benefit from load balancing techniques proposed for traditional storage systems and find only limited improvement in the measured throughput.

Disk Read-Write Optimizations and Data Integrity in Transaction Systems Using Write-Ahead Logging We discuss several disk read-write optimizations that are implemented in different transaction systems and disk hardware to improve performance. These include: (1) when multiple sectors are written to disk, the sectors may be written out of sequence (SCSI disk interfaces do this). (2) Avoiding initializing pages on disk when a file is extended. (3) Not accessing individual pages during a mass delete operation (e.g., dropping an index from a file which contains multiple indexes). (4) Permitting a previously deallocated page to be reallocated without the need to read the deallocated version of the page from disk during its reallocation. (5) Purging of file pages from the buffer pool during a file erase operation (e.g., a table drop). (6) Avoiding logging for bulk operations like index create. We consider a system which implements the above optimizations and in which a page consists of multiple disk sectors and recovery is based on write-ahead logging using a log sequence number on every page. For such a system, we present a simple method for guaranteeing the detection of the partial disk write of a page. Detecting partial writes is very important not only to ensure data integrity from the users' viewpoint but also to make the transaction system software work correctly. Once a partial write is detected, it is easy to recover such a page using media recovery techniques. Our method imposes minimal CPU and space overheads. It has been implemented in DB2/6000 and ADSM.

File grouping for scientific data management: lessons from experimenting with real traces The analysis of data usage in a large set of real traces from a high-energy physics collaboration revealed the existence of an emergent grouping of files that we coined "filecules". This paper presents the benefits of using this file grouping for prestaging data and compares it with previously proposed file grouping techniques along a range of performance metrics. Our experiments with real workloads demonstrate that filecule grouping is a reliable and useful abstraction for data management in science Grids; that preserving time locality for data prestaging is highly recommended; that job reordering with respect to data availability has significant impact on throughput; and finally, that a relatively short history of traces is a good predictor for filecule grouping. Our experimental results provide lessons for workload modeling and suggest design guidelines for data management in data-intensive resource-sharing environments.

Group-based management of distributed file caches We describe a way to manage distributed file system caches based upon groups of files that are accessed together. We use file access patterns to automatically construct dynamic groupings of files and then manage our cache by fetching groups, rather than single files. We present experimental results, based on trace-driven workloads, demonstrating that grouping improves cache performance. At the file system client, grouping can reduce LRU demand fetches by 50 to 60%. At the server cache hit rate improvements are much more pronounced, but vary widely (20 to over 1200%) depending upon the capacity of intervening caches. Our treatment includes information theoretic results that justify our approach to file grouping.

RAID: high-performance, reliable secondary storage Disk arrays were proposed in the 1980s as a way to use parallelism between multiple disks to improve aggregate I/O performance. Today they appear in the product lines of most major computer manufacturers. This article gives a comprehensive overview of disk arrays and provides a framework in which to organize current and future work. First, the article introduces disk technology and reviews the driving forces that have popularized disk arrays: performance and reliability. It discusses the two architectural techniques used in disk arrays: striping across multiple disks to improve performance and redundancy to improve reliability. Next, the article describes seven disk array architectures, called RAID (Redundant Arrays of Inexpensive Disks) levels 0–6 and compares their performance, cost, and reliability. It goes on to discuss advanced research and implementation topics such as refining the basic RAID levels to improve performance and designing algorithms to maintain data consistency. Last, the article describes six disk array prototypes of products and discusses future opportunities for research, with an annotated bibliography disk array-related literature.

An Analytic Treatment Of The Reliability And Performance Of Mirrored Disk Subsystems

Background data movement in a log-structured disk subsystem The log-structured disk subsystem is a new concept for the use of disk storage whose future application has enormous potential. In such a subsystem, all writes are organized into a log, each entry of which is placed into the next available free storage. A directory indicates the physical location of each logical object (e.g., each file block or track image) as known to the processor originating the I/O request. For those objects that have been written more than once, the directory retains the location of the most recent copy. Other work with log-structured disk subsystems has shown that they are capable of high write throughputs. However, the fragmentation of free storage due to the scattered locations of data that become out of date can become a problem in sustained operation. To control fragmentation, it is necessary to perform ongoing garbage collection, in which the location of stored data is shifted to release unused storage for re-use. This paper introduces a mathematical model of garbage collection, and shows how collection load relates to the utilization of storage and the amount of locality present in the pattern of updates. A realistic statistical model of updates, based upon trace data analysis, is applied. In addition, alternative policies are examined for determining which data areas to collect. The key conclusion of our analysis is that in environments with the scattered update patterns typical of database I/O, the utilization of storage must be controlled in order to achieve the high write throughput of which the subsystem is capable. In addition, the presence of data locality makes it important to take the past history of data into account in determining the next area of storage to be garbage-collected.

Parallel database systems: The case for shared-something Parallel database systems are becoming the primary application of multiprocessor computers. The reason for this is that they can provide high-performance and high-availability database support at a much lower price than do equivalent mainframe computers. The traditional shared-memory, shared-disk, and shared-nothing architectures of parallel database systems are compared, based on the following dimensions: simplicity, cost, performance, availability and extensibility. Based on these comparisons, the case is made for the shared-something architecture, which can provide a better trade-off between the various objectives

Context-Dependent Pre-Trained Deep Neural Networks for Large-Vocabulary Speech Recognition We propose a novel context-dependent (CD) model for large-vocabulary speech recognition (LVSR) that leverages recent advances in using deep belief networks for phone recognition. We describe a pre-trained deep neural network hidden Markov model (DNN-HMM) hybrid architecture that trains the DNN to produce a distribution over senones (tied triphone states) as its output. The deep belief network pre-training algorithm is a robust and often helpful way to initialize deep neural networks generatively that can aid in optimization and reduce generalization error. We illustrate the key components of our model, describe the procedure for applying CD-DNN-HMMs to LVSR, and analyze the effects of various modeling choices on performance. Experiments on a challenging business search dataset demonstrate that CD-DNN-HMMs can significantly outperform the conventional context-dependent Gaussian mixture model (GMM)-HMMs, with an absolute sentence accuracy improvement of 5.8% and 9.2% (or relative error reduction of 16.0% and 23.2%) over the CD-GMM-HMMs trained using the minimum phone error rate (MPE) and maximum-likelihood (ML) criteria, respectively.

Representing incomplete knowledge in abductive logic programming Recently, Gelfond and Lifschitz presented a formal language for representingincomplete knowledge on actions and states, and a sound translation from thislanguage to extended logic programming. We present an alternative translationto abductive logic programming with integrity constraints and prove thesoundness and completeness. In addition, we show how an abductive procedurecan be used, not only for explanation, but also for deduction and provingsatisfiability under uncertainty.From a...

Planning in Action Formalisms based on DLs: First Results In this paper, we continue the recently started work on inte- grating action formalisms with description logics (DLs), by investigating planning in the context of DLs. We prove that the plan existence prob- lem is decidable for actions described in fragments of ALCQIO. More precisely, we show that its computational complexity coincides with the one of projection for DLs between ALC and ALCQIO.

Learning Topic Representation For Smt With Neural Networks Statistical Machine Translation (SMT) usually utilizes contextual information to disambiguate translation candidates. However, it is often limited to contexts within sentence boundaries, hence broader topical information cannot be leveraged. In this paper, we propose a novel approach to learning topic representation for parallel data using a neural network architecture, where abundant topical contexts are embedded via topic relevant monolingual data. By associating each translation rule with the topic representation, topic relevant rules are selected according to the distributional similarity with the source text during SMT decoding. Experimental results show that our method significantly improves translation accuracy in the NIST Chinese-to-English translation task compared to a state-of-the-art baseline.

Distributed parallel data storage systems: a scalable approach to high speed image servers We have designed, built, and analyzed a distributed parallel storage system that will supply image streams fast enough to permit multi-user, “real-time”, video-like applications in a wide-area ATM network-based Internet environment. We have based the implementation on user-level code in order to secure portability; we have characterized the performance bottlenecks arising from operating system and hardware issues, and based on this have optimized our design to make the best use of the available performance. Although at this time we have only operated with a few classes of data, the approach appears to be capable of providing a scalable, high-performance, and economical mechanism to provide a data storage system for several classes of data (including mixed multimedia streams), and for applications (clients) that operate in a high-speed network environment.

Using high speed networks to enable distributed parallel image server systems We describe the design and implementation of a distributed parallel storage system that uses high-speed ATM networks as a key element of the architecture. Other elements include a collection of network-based disk block servers, and an associated name server that provides some file system functionality. The implementation is based on user level software that runs on UNIX workstations. Both the architecture and the implementation are intended to provide for easy and economical scalability. This approach has yielded a data source that scales economically to very high speed. Target applications include online storage for both very large images and video sequences. This paper describes the architecture, and explores the performance issues of the current implementation.

System issues in implementing high speed distributed parallel storage systems In this paper we describe several aspects of implementing a high speed network-based distributed application. We describe the design and implementation of a distributed parallel storage system that uses high speed ATM networks as a key element of the architecture. The architecture provides what amounts to a collection of network-based disk block servers, and an associated name server that provides some file system functionality. The implementation approach is that of user level software that runs on UNIX workstations. Both the architecture and the implementation are intended to provide for easy and economical scalability in both performance and capacity. We describe the software architecture, the implementation and operating system overhead issues, and our experiences with this approach in an IP-over-ATM WAN. Although most of the paper is specific to a distributed parallel data server, we believe many of the issues we encountered are generally applicable to any high speed network-based application.

Continuous retrieval of multimedia data using parallelism Most implementations of workstation-based multimedia information systems cannot support a continuous display of high resolution audio and video data and suffer from frequent disruptions and delays termed hiccups. This is due to the low I/O bandwidth of the current disk technology, the high bandwidth requirement of multimedia objects, and the large size of these objects, which requires them to be almost always disk resident. A parallel multimedia information system and the key technical ideas that enable it to support a real-time display of multimedia objects are described. In this system, a multimedia object across several disk drives is declustered, enabling the system to utilize the aggregate bandwidth of multiple disks to retrieve an object in real-time. Then, the workload of an application is distributed evenly across the disk drives to maximize the processing capability of the system. To support simultaneous display of several multimedia objects for different users, two alternative approaches are described. The first approach multitasks a disk drive among several requests while the second replicates the data and dedicates resources to each individual request. The trade-offs associated with each approach are investigated using a simulation model.

A case for redundant arrays of inexpensive disks (RAID) Increasing performance of CPUs and memories will be squandered if not matched by a similar performance increase in I/O. While the capacity of Single Large Expensive Disks (SLED) has grown rapidly, the performance improvement of SLED has been modest. Redundant Arrays of Inexpensive Disks (RAID), based on the magnetic disk technology developed for personal computers, offers an attractive alternative to SLED, promising improvements of an order of magnitude in performance, reliability, power consumption, and scalability. This paper introduces five levels of RAIDs, giving their relative cost/performance, and compares RAID to an IBM 3380 and a Fujitsu Super Eagle.

Generalized working sets for segment reference strings The working-set concept is extended for programs that reference segments of different sizes. The generalized working-set policy (GWS) keeps as its resident set those segments whose retention costs do not exceed their retrieval costs. The GWS is a model for the entire class of demand-fetching memory policies that satisfy a resident-set inclusion property. A generalized optimal policy (GOPT) is also defined; at its operating points it minimizes aggregated retention and swapping costs. Special cases of the cost structure allow GWS and GOPT to simulate any known stack algorithm, the working set, and VMIN. Efficient procedures for computing demand curves showing swapping load as a function of memory usage are developed for GWS and GOPT policies. Empirical data from an actual system are included.

A combined method for maintaining large indices in multiprocessor multidisk environments Consider the problem of maintaining large indices (or secondary memory indices) in a multiprocessor multidisk environment in which each processor has a dedicated secondary memory (one disk or more). The processors either reside in the same site and communicate via shared memory, or reside in different sites and communicate via a local broadcast network. The straightforward method (SFM) for maintaining such an index, which is commonly called declustering, is to partition the index records equally among the processors, each of which maintains its part of the index in a local B/sup tree. In prior work (Inform. Processing Lett., vol. 34, pp. 313-321, May 1990), we have presented another method, called the "totally distributed B/sup tree" (TDB) method, in which all processors together implement a "wide" B/sup tree. There are settings in which the second method is better than the first method, and vice versa. In this paper, we present a new method, called the combined distribution method (CDM), that combines the ideas underlying SFM and TDB. In tightly coupled environments, CDM outperforms both SFM and TDB in almost all practical settings (in many settings by more than 30%). This is shown by an approximate analysis and verified by simulations. Note that CDM's approach can improve performance in database systems that use a RAID (redundant array of inexpensive disks).

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Integrated document caching and prefetching in storage hierarchies based on Markov-chain predictions Large multimedia document archives may hold a major fraction of their data in tertiary storage libraries for cost reasons. This paper develops an integrated approach to the vertical data migration between the tertiary, secondary, and primary storage in that it reconciles speculative prefetching, to mask the high latency of the tertiary storage, with the replacement policy of the document caches at the secondary and primary storage level, and also considers the interaction of these policies with the tertiary and secondary storage request scheduling. The integrated migration policy is based on a continuous-time Markov chain model for predicting the expected number of accesses to a document within a specified time horizon. Prefetching is initiated only if that expectation is higher than those of the documents that need to be dropped from secondary storage to free up the necessary space. In addition, the possible resource contention at the tertiary and secondary storage is taken into account by dynamically assessing the response-time benefit of prefetching a document versus the penalty that it would incur on the response time of the pending document requests. The parameters of the continuous-time Markov chain model, the probabilities of co-accessing certain documents and the interaction times between successive accesses, are dynamically estimated and adjusted to evolving workload patterns by keeping online statistics. The integrated policy for vertical data migration has been implemented in a prototype system. The system makes profitable use of the Markov chain model also for the scheduling of volume exchanges in the tertiary storage library. Detailed simulation experiments with Web-server-like synthetic workloads indicate significant gains in terms of client response time. The experiments also show that the overhead of the statistical bookkeeping and the computations for the access predictions is affordable.

A theory of diagnosis from first principles Without Abstract

Query Order We study the effect of query order on computational power and show that ${\rm P}^{{\rm BH}_j[1]:{\rm BH}_k[1]}$\allowbreak---the languages computable via a polynomial-time machine given one query to the $j$th level of the boolean hierarchy followed by one query to the $k$th level of the boolean hierarchy---equals ${\rm R}_{{j+2k-1}{\scriptsize\mbox{-tt}}}^{p}({\rm NP})$ if $j$ is even and $k$ is odd and equals ${\rm R}_{{j+2k}{\scriptsize\mbox{-tt}}}^{p}({\rm NP})$ otherwise. Thus unless the polynomial hierarchy collapses it holds that, for each $1\leq j \leq k$, ${\rm P}^{{\rm BH}_j[1]:{\rm BH}_k[1]} = {\rm P}^{{\rm BH}_k[1]:{\rm BH}_j [1]} \iff (j=k) \lor (j\mbox{ is even}\, \land k=j+1)$. We extend our analysis to apply to more general query classes.

Complexity Results for Planning I describe several computational complexity results for planning, some of which identify tractable planning problems. The model of planning, called "propositional planning," is simple—conditions within operators are literals with no variables allowed. The different plan­ ning problems are defined by different restric- tions on the preconditions and postconditions of operators. The main results are: Proposi­ tional planning is PSPACE-complete, even if operators are restricted to two positive (non- negated) preconditions and two postconditions, or if operators are restricted to one postcondi­ tion (with any number of preconditions ). It is NP-complete if operators are restricted to positive postconditions, even if operators are restricted to one precondition and one posi­ tive postcondition. It is tractable in a few re­ stricted cases, one of which is if each opera­ tor is restricted to positive preconditions and one postcondition. The blocks-world problem, slightly modified, is a subproblem of this re­ stricted planning problem.

Improving the tolerance of multilayer perceptrons by minimizing the statistical sensitivity to weight deviations This paper proposes a version of the backpropagation algorithm which increases the tolerance of a feedforward neural network against deviations in the weight values. These changes can originate either when the neural network is mapped on a given VLSI circuit where the precision and/or weight matching are low, or by physical defects affecting the neural circuits. The modified backpropagation algorithm we propose uses the statistical sensitivity of the network to changes in the weights as a quantitative measure of network tolerance and attempts to reduce this statistical sensitivity while keeping the figures for the usual training performance (in errors and time) similar to those obtained with the usual backpropagation algorithm.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Knowledge Matters: Importance of Prior Information for Optimization We explored the effect of introducing prior knowledge into the intermediate level of deep supervised neural networks on two tasks. On a task we designed, all black-box state-of-theart machine learning algorithms which we tested, failed to generalize well. We motivate our work from the hypothesis that, there is a training barrier involved in the nature of such tasks, and that humans learn useful intermediate concepts from other individuals by using a form of supervision or guidance using a curriculum. Our results provide a positive evidence in favor of this hypothesis. In our experiments, we trained a two-tiered MLP architecture on a dataset for which each input image contains three sprites, and the binary target class is 1 if all of three shapes belong to the same category and otherwise the class is 0. In terms of generalization, black-box machine learning algorithms could not perform better than chance on this task. Standard deep supervised neural networks also failed to generalize. However, using a particular structure and guiding the learner by providing intermediate targets in the form of intermediate concepts (the presence of each object) allowed us to solve the task efficiently. We obtained much better than chance, but imperfect results by exploring different architectures and optimization variants. This observation might be an indication of optimization difficulty when the neural network trained without hints on this task. We hypothesize that the learning difficulty is due to the composition of two highly non-linear tasks. Our findings are also consistent with the hypotheses on cultural learning inspired by the observations of training of neural networks sometimes getting stuck, even though good solutions exist, both in terms of training and generalization error.

Time Series Compression Based on Adaptive Piecewise Recurrent Autoencoder. Time series account for a large proportion of the data stored in financial, medical and scientific databases. The efficient storage of time series is important in practical applications. In this paper, we propose a novel compression scheme for time series. The encoder and decoder are both composed by recurrent neural networks (RNN) such as long short-term memory (LSTM). There is an autoencoder between encoder and decoder, which encodes the hidden state and input together and decodes them at the decoder side. Moreover, we pre-process the original time series by partitioning it into segments with various lengths which have similar total variation. The experimental study shows that the proposed algorithm can achieve competitive compression ratio on real-world time series.

Lightweight Lossy Compression of Biometric Patterns via Denoising Autoencoders Wearable Internet of Things (IoT) devices permit the massive collection of biosignals (e.g., heart-rate, oxygen level, respiration, blood pressure, photo-plethysmographic signal, etc.) at low cost. These, can be used to help address the individual fitness needs of the users and could be exploited within personalized healthcare plans. In this letter, we are concerned with the design of lightweight and efficient algorithms for the lossy compression of these signals. In fact, we underline that compression is a key functionality to improve the lifetime of IoT devices, which are often energy constrained, allowing the optimization of their internal memory space and the efficient transmission of data over their wireless interface. To this end, we advocate the use of autoencoders as an efficient and computationally lightweight means to compress biometric signals. While the presented techniques can be used with any signal showing a certain degree of periodicity, in this letter we apply them to ECG traces, showing quantitative results in terms of compression ratio, reconstruction error and computational complexity. State of the art solutions are also compared with our approach.

On building ensembles of stacked denoising auto-encoding classifiers and their further improvement. •Explores how binarization permits/improves diversification in deep machines.•Shows the effectiveness of pre-emphasizing samples for deep classification.•Combines the above with data augmentation to reach record results.•Opens further research lines in deep learning.

What regularized auto-encoders learn from the data-generating distribution. What do auto-encoders learn about the underlying data-generating distribution? Recent work suggests that some auto-encoder variants do a good job of capturing the local manifold structure of data. This paper clarifies some of these previous observations by showing that minimizing a particular form of regularized reconstruction error yields a reconstruction function that locally characterizes the shape of the data-generating density. We show that the auto-encoder captures the score (derivative of the log-density with respect to the input). It contradicts previous interpretations of reconstruction error as an energy function. Unlike previous results, the theorems provided here are completely generic and do not depend on the parameterization of the auto-encoder: they show what the auto-encoder would tend to if given enough capacity and examples. These results are for a contractive training criterion we show to be similar to the denoising auto-encoder training criterion with small corruption noise, but with contraction applied on the whole reconstruction function rather than just encoder. Similarly to score matching, one can consider the proposed training criterion as a convenient alternative to maximum likelihood because it does not involve a partition function. Finally, we show how an approximate Metropolis-Hastings MCMC can be setup to recover samples from the estimated distribution, and this is confirmed in sampling experiments.

A Deep and Tractable Density Estimator. The Neural Autoregressive Distribution Estimator (NADE) and its real-valued version RNADE are competitive density models of multidimensional data across a variety of domains. These models use a fixed, arbitrary ordering of the data dimensions. One can easily condition on variables at the beginning of the ordering, and marginalize out variables at the end of the ordering, however other inference tasks require approximate inference. In this work we introduce an efficient procedure to simultaneously train a NADE model for each possible ordering of the variables, by sharing parameters across all these models. We can thus use the most convenient model for each inference task at hand, and ensembles of such models with different orderings are immediately available. Moreover, unlike the original NADE, our training procedure scales to deep models. Empirically, ensembles of Deep NADE models obtain state of the art density estimation performance.

Deep Sparse Rectifier Neural Networks.

Understanding the difficulty of training deep feedforward neural networks Whereas before 2006 it appears that deep multilayer neural networks were not successfully trained, since then several algorithms have been shown to successfully train them, with experimental results showing the superiority of deeper vs less deep architectures. All these experimental results were obtained with new initialization or training mechanisms. Our objective here is to understand better why standard gradient descent from random initialization is doing so poorly with deep neural networks, to better understand these recent relative successes and help design better algorithms in the future. We first observe the influence of the non-linear activations functions. We find that the logistic sigmoid activation is unsuited for deep networks with random initialization because of its mean value, which can drive especially the top hidden layer into saturation. Surprisingly, we find that saturated units can move out of saturation by themselves, albeit slowly, and explaining the plateaus sometimes seen when training neural networks. We find that a new non-linearity that saturates less can often be beneficial. Finally, we study how activations and gradients vary across layers and during training, with the idea that training may be more difficult when the singular values of the Jacobian associated with each layer are far from 1. Based on these considerations, we propose a new initialization scheme that brings substantially faster convergence. 1 Deep Neural Networks Deep learning methods aim at learning feature hierarchies with features from higher levels of the hierarchy formed by the composition of lower level features. They include Appearing in Proceedings of the 13 International Conference on Artificial Intelligence and Statistics (AISTATS) 2010, Chia Laguna Resort, Sardinia, Italy. Volume 9 of JMLR: WC Weston et al., 2008). Much attention has recently been devoted to them (see (Bengio, 2009) for a review), because of their theoretical appeal, inspiration from biology and human cognition, and because of empirical success in vision (Ranzato et al., 2007; Larochelle et al., 2007; Vincent et al., 2008) and natural language processing (NLP) (Collobert \u0026 Weston, 2008; Mnih \u0026 Hinton, 2009). Theoretical results reviewed and discussed by Bengio (2009), suggest that in order to learn the kind of complicated functions that can represent high-level abstractions (e.g. in vision, language, and other AI-level tasks), one may need deep architectures. Most of the recent experimental results with deep architecture are obtained with models that can be turned into deep supervised neural networks, but with initialization or training schemes different from the classical feedforward neural networks (Rumelhart et al., 1986). Why are these new algorithms working so much better than the standard random initialization and gradient-based optimization of a supervised training criterion? Part of the answer may be found in recent analyses of the effect of unsupervised pretraining (Erhan et al., 2009), showing that it acts as a regularizer that initializes the parameters in a “better” basin of attraction of the optimization procedure, corresponding to an apparent local minimum associated with better generalization. But earlier work (Bengio et al., 2007) had shown that even a purely supervised but greedy layer-wise procedure would give better results. So here instead of focusing on what unsupervised pre-training or semi-supervised criteria bring to deep architectures, we focus on analyzing what may be going wrong with good old (but deep) multilayer neural networks. Our analysis is driven by investigative experiments to monitor activations (watching for saturation of hidden units) and gradients, across layers and across training iterations. We also evaluate the effects on these of choices of activation function (with the idea that it might affect saturation) and initialization procedure (since unsupervised pretraining is a particular form of initialization and it has a drastic impact).

Three new graphical models for statistical language modelling The supremacy of n-gram models in statistical language modelling has recently been challenged by parametric models that use distributed representations to counteract the difficulties caused by data sparsity. We propose three new probabilistic language models that define the distribution of the next word in a sequence given several preceding words by using distributed representations of those words. We show how real-valued distributed representations for words can be learned at the same time as learning a large set of stochastic binary hidden features that are used to predict the distributed representation of the next word from previous distributed representations. Adding connections from the previous states of the binary hidden features improves performance as does adding direct connections between the real-valued distributed representations. One of our models significantly outperforms the very best n-gram models.

Links between perceptrons, MLPs and SVMs We propose to study links between three important classification algorithms: Perceptrons, Multi-Layer Perceptrons (MLPs) and Support Vector Machines (SVMs). We first study ways to control the capacity of Perceptrons (mainly regularization parameters and early stopping), using the margin idea introduced with SVMs. After showing that under simple conditions a Perceptron is equivalent to an SVM, we show it can be computationally expensive in time to train an SVM (and thus a Perceptron) with stochastic gradient descent, mainly because of the margin maximization term in the cost function. We then show that if we remove this margin maximization term, the learning rate or the use of early stopping can still control the margin. These ideas are extended afterward to the case of MLPs. Moreover, under some assumptions it also appears that MLPs are a kind of mixture of SVMs, maximizing the margin in the hidden layer space. Finally, we present a very simple MLP based on the previous findings, which yields better performances in generalization and speed than the other models.

Distributing a B+-tree in a loosely coupled environment We consider the problem of maintaining a data file which must be distributed among disks, each controlled by a processor and residing in a different site. The pairs, of disks and processors, are connected via a local broadcast network. A simple and practical method is presented.

Ignoring Irrelevant Facts and Operators in Plan Generation . It is traditional wisdom that one should start from the goalswhen generating a plan in order to focus the plan generation process onpotentially relevant actions. The graphplan system, however, which isthe most efficient planning system nowadays, builds a &quot;planning graph&quot;in a forward-chaining manner. Although this strategy seems to workwell, it may possibly lead to problems if the planning task descriptioncontains irrelevant information. Although some irrelevant informationcan be ...

Finding a Shortest Solution for the N × N Extension of the 15-PUZZLE Is Intractable

Deep Belief Network-Based Approaches for Link Prediction in Signed Social Networks In some online social network services (SNSs), the members are allowed to label their relationships with others, and such relationships can be represented as the links with signed values (positive or negative). The networks containing such relations are named signed social networks (SSNs), and some real-world complex systems can be also modeled with SSNs. Given the information of the observed structure of an SSN, the link prediction aims to estimate the values of the unobserved links. Noticing that most of the previous approaches for link prediction are based on the members' similarity and the supervised learning method, however, research work on the investigation of the hidden principles that drive the behaviors of social members are rarely conducted. In this paper, the deep belief network (DBN)-based approaches for link prediction are proposed. Including an unsupervised link prediction model, a feature representation method and a DBN-based link prediction method are introduced. The experiments are done on the datasets from three SNSs (social networking services) in different domains, and the results show that our methods can predict the values of the links with high performance and have a good generalization ability across these datasets.

Constraint Logic Programming for Local and Symbolic Model-Checking We propose a model checking scheme for a semantically complete fragment of CTL by combining techniques from constraint logic programming, a restricted form of constructive negation and tabled resolution. Our approach is symbolic in that it encodes and manipulates sets of states using constraints; it supports local model checking using goal-directed computation enhanced by tabulation. The framework is parameterized by the constraint domain and supports any finite constraint domain closed under disjunction, projection and complementation. We show how to encode our fragment of CTL in constraint logic programming; we outline an abstract execution model for the resulting type of programs and provide a preliminary evaluation of the approach.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

Parameterized complexity for the database theorist

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Training Hierarchical Feed-Forward Visual Recognition Models Using Transfer Learning from Pseudo-Tasks Building visual recognition models that adapt across different domains is a challenging task for computer vision. While feature-learning machines in the form of hierarchial feed-forward models (e.g., convolutional neural networks) showed promise in this direction, they are still difficult to train especially when few training examples are available. In this paper, we present a framework for training hierarchical feed-forward models for visual recognition, using transfer learning from pseudo tasks. These pseudo tasks are automatically constructed from data without supervision and comprise a set of simple pattern-matching operations. We show that these pseudo tasks induce an informative inverse-Wishart prior on the functional behavior of the network, offering an effective way to incorporate useful prior knowledge into the network training. In addition to being extremely simple to implement, and adaptable across different domains with little or no extra tuning, our approach achieves promising results on challenging visual recognition tasks, including object recognition, gender recognition, and ethnicity recognition.

The C-loss function for pattern classification This paper presents a new loss function for neural network classification, inspired by the recently proposed similarity measure called Correntropy. We show that this function essentially behaves like the conventional square loss for samples that are well within the decision boundary and have small errors, and L"0 or counting norm for samples that are outliers or are difficult to classify. Depending on the value of the kernel size parameter, the proposed loss function moves smoothly from convex to non-convex and becomes a close approximation to the misclassification loss (ideal 0-1 loss). We show that the discriminant function obtained by optimizing the proposed loss function in the neighborhood of the ideal 0-1 loss function to train a neural network is immune to overfitting, more robust to outliers, and has consistent and better generalization performance as compared to other commonly used loss functions, even after prolonged training. The results also show that it is a close competitor to the SVM. Since the proposed method is compatible with simple gradient based online learning, it is a practical way of improving the performance of neural network classifiers.

A loss function for classification based on a robust similarity metric We present a margin-based loss function for classification, inspired by the recently proposed similarity measure called correntropy. We show that correntropy induces a nonconvex loss function that is a closer approximation to the misclassification loss (ideal 0-1 loss). We show that the discriminant function obtained by optimizing the proposed loss function using a neural network is insensitive to outliers and has better generalization performance as compared to using the squared loss function which is common in neural network classifiers. The proposed method of training classifiers is a practical way of obtaining better results on real world classification problems, that uses a simple gradient based online training procedure for minimizing the empirical risk.

An Information Measure For Classification

Training connectionist models for the structured language model We investigate the performance of the Structured Language Model (SLM) in terms of perplexity (PPL) when its components are modeled by connectionist models. The connectionist models use a distributed representation of the items in the history and make much better use of contexts than currently used interpolated or back-off models, not only because of the inherent capability of the connectionist model in fighting the data sparseness problem, but also because of the sublinear growth in the model size when the context length is increased. The connectionist models can be further trained by an EM procedure, similar to the previously used procedure for training the SLM. Our experiments show that the connectionist models can significantly improve the PPL over the interpolated and back-off models on the UPENN Treebank corpora, after interpolating with a baseline trigram language model. The EM training procedure can improve the connectionist models further, by using hidden events obtained by the SLM parser.

Nonlocal estimation of manifold structure. We claim and present arguments to the effect that a large class of manifold learning algorithms that are essentially local and can be framed as kernel learning algorithms will suffer from the curse of dimensionality, at the dimension of the true underlying manifold. This observation invites an exploration of nonlocal manifold learning algorithms that attempt to discover shared structure in the tangent planes at different positions. A training criterion for such an algorithm is proposed, and experiments estimating a tangent plane prediction function are presented, showing its advantages with respect to local manifold learning algorithms: it is able to generalize very far from training data (on learning handwritten character image rotations), where local nonparametric methods fail.

Unsupervised Learning of Models for Recognition We present a method to learn object class models from unlabeled and unsegmented cluttered scenes for the purpose of visual object recognition. We focus on a particular type of model where objects are represented as flexible constellations of rigid parts (features). The variability within a class is represented by a joint probability density function (pdf) on the shape of the constellation and the output of part detectors. In a first stage, the method automatically identifies distinctive parts in the training set by applying a clustering algorithm to patterns selected by an interest operator. It then learns the statistical shape model using expectation maximization. The method achieves very good classification results on human faces and rear views of cars.

Using Manifold Stucture for Partially Labeled Classification Abstract: We consider the general problem of utilizing both labeled and unlabeleddata to improve classification accuracy. Under the assumptionthat the data lie on a submanifold in a high dimensional space,we develop an algorithmic framework to classify a partially labeleddata set in a principled manner. The central idea of our approach isthat classification functions are naturally defined only on the submanifoldin question rather than the total ambient space. Using theLaplace Beltrami...

Sparse Feature Learning for Deep Belief Networks Unsupervised learning algorithms aim to discover the structure hidden in the data, and to learn representations that are more suitable as input to a supervised machine than the raw input. Many unsupervised methods are based on reconstructing the input from the representation, while constraining the repr esentation to have cer- tain desirable properties (e.g. low dimension, sparsity, e tc). Others are based on approximating density by stochastically reconstructing t he input from the repre- sentation. We describe a novel and efficient algorithm to lea rn sparse represen- tations, and compare it theoretically and experimentally with a similar machine trained probabilistically, namely a Restricted Boltzmann Machine. We propose a simple criterion to compare and select different unsupervised machines based on the trade-off between the reconstruction error and the info rmation content of the representation. We demonstrate this method by extracting features from a dataset of handwritten numerals, and from a dataset of natural image patches. We show that by stacking multiple levels of such machines and by training sequentially, high-order dependencies between the input observed variables can be captured.

Advances in optimizing recurrent networks After a more than decade-long period of relatively little research activity in the area of recurrent neural networks, several new developments will be reviewed here that have allowed substantial progress both in understanding and in technical solutions towards more efficient training of recurrent networks. These advances have been motivated by and related to the optimization issues surrounding deep learning. Although recurrent networks are extremely powerful in what they can in principle represent in terms of modeling sequences, their training is plagued by two aspects of the same issue regarding the learning of long-term dependencies. Experiments reported here evaluate the use of clipping gradients, spanning longer time ranges with leaky integration, advanced momentum techniques, using more powerful output probability models, and encouraging sparser gradients to help symmetry breaking and credit assignment. The experiments are performed on text and music data and show off the combined effects of these techniques in generally improving both training and test error.

Multimodal fusion using dynamic hybrid models We propose a novel hybrid model that exploits the strength of discriminative classifiers along with the representational power of generative models. Our focus is on detecting multimodal events in time varying sequences. Discriminative classifiers have been shown to achieve higher performances than the corresponding generative likelihood-based classifiers. On the other hand, generative models learn a rich informative space which allows for data generation and joint feature representation that discriminative models lack. We employ a deep temporal generative model for unsupervised learning of a shared representation across multiple modalities with time varying data. The temporal generative model takes into account short term temporal phenomena and allows for filling in missing data by generating data within or across modalities. The hybrid model involves augmenting the temporal generative model with a temporal discriminative model for event detection, and classification, which enables modeling long range temporal dynamics. We evaluate our approach on audio-visual datasets (AVEC, AVLetters, and CUAVE) and demonstrate its superiority compared to the state-of-the-art.

A Goal-Oriented Approach to Computing Well Founded Semantics

Planning with sensing, concurrency, and exogenous events: logical framework and implementation The focus of current research in cognitive robotics is both on the realization of sys- tems based on known formal settings and on the extension of previous formal approaches to account for features that play a signifl- cant role for autonomous robots, but have not yet received an adequate treatment. In this paper we adopt a formal framework de- rived from Propositional Dynamic Logics by exploiting their formal correspondence with Description Logics, and present an extension of such a framework obtained by introducing both concurrency on primitive actions and autoepistemic operators for explicitly repre- senting the robot's epistemic state. We show that the resulting formal setting allows for the representation of actions with context- dependent efiects, sensing actions, and con- current actions, and address both the pres- ence of exogenous events and the characteri- zation of the notion of executable plan in such a complex setting. Moreover, we present an implementation of this framework in a system which is capable of generating plans that are actually executed on mobile robots, and illus- trate the experimentation of such a system in the design and implementation of soccer players for the 1999 Robocup competition.

Editorial introduction to the Neural Networks special issue on Deep Learning of Representations.

TCP Nice: a mechanism for background transfers Many distributed applications can make use of large background transfers--transfers of data that humans are not waiting for--to improve availability, reliability, latency or consistency. However, given the rapid fluctuations of available network bandwidth and changing resource costs due to technology trends, hand tuning the aggressiveness of background transfers risks (1) complicating applications, (2) being too aggressive and interfering with other applications, and (3) being too timid and not gaining the benefits of background transfers. Our goal is for the operating system to manage network resources in order to provide a simple abstraction of near zero-cost background transfers. Our system, TCP Nice, can provably bound the interference inflicted by background flows on foreground flows in a restricted network model. And our microbenchmarks and case study applications suggest that in practice it interferes little with foreground flows, reaps a large fraction of spare network bandwidth, and simplifies application construction and deployment. For example, in our prefetching case study application, aggressive prefetching improves demand performance by a factor of three when Nice manages resources; but the same prefetching hurts demand performance by a factor of six under standard network congestion control.

PI/OT: parallel I/O templates This paper presents a novel, top-down, high-level approach to parallelizing file I/O. Each parallel file descriptor is annotated with a high-level specification, or template, of the expected parallel behavior. The annotations are external to and independent of the source code. At run-time, all I/O using a parallel file descriptor adheres to the semantics of the selected template. By separating the parallel I/O specifications from the code, a user can quickly change the I/O behavior without rewriting the code. Templates can be composed hierarchically to construct complex access patterns. Two sample parallel programs using these templates are compared against versions implemented in an existing parallel I/O system (PIOUS). The sample programs show that the use of parallel I/O templates are beneficial from both the performance and software engineering points of view.

Latency management in storage systems Storage Latency Estimation Descriptors, or SLEDs, are an API that allow applications to understand and take advantage of the dynamic state of a storage system. By accessing data in the file system cache or high-speed storage first, total I/O workloads can be reduced and performance improved. SLEDs report estimated data latency, allowing users, system utilities, and scripts to make file access decisions based on those retrieval time estimates. SLEDs thus can be used to improve individual application performance, reduce system workloads, and improve the user experience with more predictable behavior. We have modified the Linux 2.2 kernel to support SLEDs, and several Unix utilities and astronomical applications have been modified to use them. As a result, execution times of the Unix utilities when data file sizes exceed the size of the file system buffer cache have been reduced from 50% up to more than an order of magnitude. The astronomical applications incurred 30-50% fewer page faults and reductions in execution time of 10-35%. Performance of applications which use SLEDs also degrade more gracefully as data file size grows.

Exploiting the non-determinism and asynchrony of set iterators to reduce aggregate file I/O latency A key goal of distributed systems is to provide prompt access to shared information repositories. The high latency of remote access is a serious impediment to this goal. This paper describes a new file system abstraction called dynamic sets - unordered collections created by an application to hold the files it intends to process. Applications that iterate on the set to access its members allow the system to reduce the aggregateU0 Iatency by exploiting the non-determinism and asychrony inherent in the semantics of set iterators. This reduction in latency comes without relying on reference locality, without modifying DFS servers and protocols, and without unduly complicating the programming model. This paperpresents this abstraction and describes an implementation of it that runs on local and distributedfile systems, as well as the World wide Web. Dynamicsets demonstrate substantial performance gains - up to 50% savings in runtbne for search on NFS, and up to 90% reduction in I/O latency for Web searches.

Dealing with disaster: surviving misbehaved kernel extensions Today's extensible operating systems allow applications to modify kernel behavior by providing mechanisms for application code to run in the kernel address space. The advantage of this approach is that it provides improved application flexibility and performance; the disadvan- tage is that buggy or malicious code can jeopardize the integrity of the kernel. It has been demonstrated that it is feasible to use safe languages, software fault isolation, or virtual memory protection to safeguard the main ker- nel. However, such protection mechanisms do not address the full range of problems, such as resource hoarding, that can arise when application code is intro- duced into the kernel. In this paper, we present an analysis of extension mechanisms in the VINO kernel. VINO uses software fault isolation as its safety mechanism and a lightweight transaction system to cope with resource-hoarding. We explain how these two mechanisms are sufficient to protect against a large class of errant or malicious extensions, and we quantify the overhead that this protection introduces. We find that while the overhead of these techniques is high relative to the cost of the extensions themselves, it is low relative to the benefits that extensibility brings.

Freeblock Scheduling Outside of Disk Firmware Freeblock scheduling replaces a disk drive's rotational latency delays with useful background media transfers, potentially allowing background disk I/O to occur with no impact on foreground service times. To do so, a freeblock scheduler must be able to very accurately predict the service time components of any given disk request - the necessary accuracy was not previously considered achievable outside of disk firmware. This paper describes the design and implementation of a working external freeblock scheduler running either as a user-level application atop Linux or inside the FreeBSD kernel. This freeblock scheduler can give 15% of a disk's potential bandwidth (over 3.1MB/s) to a background disk scanning task with almost no impact (less than 2%) on the foreground request response times. This can increase disk bandwidth utilization by over 6 x.

ELFSR0: object-oriented extensible file systems High performance scientific data analysis is plagued by chronically inadequate I/O performance.The situation is aggravated by ever improving processor performance. For high performancemulticomputers, such as the Touchstone Delta that possess in excess of 500, 60 megaflops,processor I/O will be the bottleneck for many scientific applications.This report describes ELFS (an ExtensibLe File System). ELFS attacks the problems of 1)providing high bandwidth and low latency I/O to applications...

The design of POSTGRES This paper presents the preliminary design of a new database management system, called POSTGRES, that is the successor to the INGRES relational database system. The main design goals of the new system are toprovide better support for complex objects,provide user extendibility for data types, operators and access methods,provide facilities for active databases (i.e., alerters and triggers) and inferencing including forward- and backward-chaining,simplify the DBMS code for crash recovery,produce a design that can take advantage of optical disks, workstations composed of multiple tightly-coupled processors, and custom designed VLSI chips, andmake as few changes as possible (preferably none) to the relational model.The paper describes the query language, programming language interface, system architecture, query processing strategy, and storage system for the new system.

ARIMA time series modeling and forecasting for adaptive I/O prefetching Bursty application I/O patterns, together with transfer limited storage devices, combine to create a major I/O bottleneck on parallel systems. This paper explores the use of time series models to forecast application I/O request times, then prefetching I/O requests during computation intervals to hide I/O latency. Experimental results with I/O intensive scientific codes show performance improvements compared to standard UNIX prefetching strategies.

Parallel I/O Performance Characterization of Columbia and NEC SX-8 Superclusters Many scientific applications running on today's supercomputers deal with increasingly large data sets and are correspondingly bottlenecked by the time it takes to read or write the data from/to the file system. We therefore undertook a study to characterize the parallel I/O performance of two of today's leading parallel supercomputers: the Columbia system at NASA Ames Research Center and the NEC SX-8 supercluster at the University of Stuttgart, Germany. On both systems, we ran a total of seven parallel I/O benchmarks, comprising five low-level benchmarks: (i) IO_Bench, (ii) MPI Tile IO, (iii) IOR (POSIX and MPI-IO), (iv) b_eff_io (five different patterns), and (v) SPIOBENCH, and two scalable synthetic compact application (SSCA) benchmarks: (a) HPCS (High Productivity Computing Systems) SSCA #3 and (b) FLASH IO (parallel HDF5). We present the results of these experiments characterizing the parallel I/O performance of these two systems.

On Variable Scope of Parity Protection in Disk Arrays In a common form of a RAID 5 architecture, data is organized on a disk array consisting of N + 1 disks into stripes of N data blocks and one parity block (with parity block locations staggered so as to balance the number of parity blocks on each disk). This allows data to be recovered in the event of a single disk failure. Here we consider an extension to this architecture in which parity information applies to arbitrary subsets of the data blocks in each stripe. Using several simplifying assumptions, we present simulation and analytic results that provide estimates of the improvement using this approach, in terms of total I/O operations, as compared to 1) conventional RAID 5 under a random single-block write workload, and 2) the use of a log-structured file system in which data is written out in stripes. Results on the reduction of disk recovery costs are also presented.

Accuracy of admissible heuristic functions in selected planning domains The efficiency of optimal planning algorithms based on heuristic search crucially depends on the accuracy of the heuristic function used to guide the search. Often, we are interested in domain-independent heuristics for planning. In order to assess the limitations of domain-independent heuristic planning, we analyze the (in)accuracy of common domain-independent planning heuristics in the IPC benchmark domains. For a selection of these domains, we analytically investigate the accuracy of the h+ heuristic, the hm family of heuristics, and certain (additive) pattern database heuristics, compared to the perfect heuristic h*. Whereas h+and additive pattern database heuristics usually return cost estimates proportional to the true cost, non-additive hm and nonadditive pattern-database heuristics can yield results underestimating the true cost by arbitrarily large factors.

De-indirection for flash-based SSDs with nameless writes We present Nameless Writes, a new device interface that removes the need for indirection in modern solid-state storage devices (SSDs). Nameless writes allow the device to choose the location of a write; only then is the client informed of the name (i.e., address) where the block now resides. Doing so allows the device to control block-allocation decisions, thus enabling it to execute critical tasks such as garbage collection and wear leveling, while removing the need for large and costly indirection tables. We demonstrate the effectiveness of nameless writes by porting the Linux ext3 file system to use an emulated nameless-writing device and show that doing so both reduces space and time overheads, thus making for simpler, less costly, and higher-performance SSD-based storage.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

UOW-SHEF: SimpLex -- lexical simplicity ranking based on contextual and psycholinguistic features This paper describes SimpLex, a Lexical Simplification system that participated in the English Lexical Simplification shared task at SemEval-2012. It operates on the basis of a linear weighted ranking function composed of context sensitive and psycholinguistic features. The system outperforms a very strong baseline, and ranked first on the shared task.

Faster and smaller N-gram language models N-gram language models are a major resource bottleneck in machine translation. In this paper, we present several language model implementations that are both highly compact and fast to query. Our fastest implementation is as fast as the widely used SRILM while requiring only 25% of the storage. Our most compact representation can store all 4 billion n-grams and associated counts for the Google n-gram corpus in 23 bits per n-gram, the most compact lossless representation to date, and even more compact than recent lossy compression techniques. We also discuss techniques for improving query speed during decoding, including a simple but novel language model caching technique that improves the query speed of our language models (and SRILM) by up to 300%.

Out in the Open: Finding and Categorising Errors in the Lexical Simplification Pipeline. Lexical simplification is the task of automatically reducing the complexity of a text by identifying difficult words and replacing them with simpler alternatives. Whilst this is a valuable application of natural language generation, rudimentary lexical simplification systems suffer from a high error rate which often results in nonsensical, non-simple text. This paper seeks to characterise and quantify the errors which occur in a typical baseline lexical simplification system. We expose 6 distinct categories of error and propose a classification scheme for these. We also quantify these errors for a moderate size corpus, showing the magnitude of each error type. We find that for 183 identified simplification instances, only 19 (10.38%) result in a valid simplification, with the rest causing errors of varying gravity.

Putting it simply: a context-aware approach to lexical simplification We present a method for lexical simplification. Simplification rules are learned from a comparable corpus, and the rules are applied in a context-aware fashion to input sentences. Our method is unsupervised. Furthermore, it does not require any alignment or correspondence among the complex and simple corpora. We evaluate the simplification according to three criteria: preservation of grammaticality, preservation of meaning, and degree of simplification. Results show that our method outperforms an established simplification baseline for both meaning preservation and simplification, while maintaining a high level of grammaticality.

More accurate tests for the statistical significance of result differences Statistical significance testing of differences in values of metrics like recall, precision and balanced F-score is a necessary part of empirical natural language processing. Unfortunately, we find in a set of experiments that many commonly used tests often underestimate the significance and so are less likely to detect differences that exist between different techniques. This underestimation comes from an independence assumption that is often violated. We point out some useful tests that do not make this assumption, including computationally-intensive randomization tests.

Indexing By Latent Semantic Analysis

A Nonparametric Bayesian Approach to Modeling Overlapping Clusters Although clustering data into mutually ex- clusive partitions has been an extremely suc- cessful approach to unsupervised learning, there are many situations in which a richer model is needed to fully represent the data. This is the case in problems where data points actually simultaneously belong to mul- tiple, overlapping clusters. For example a particular gene may have several functions, therefore belonging to several distinct clus- ters of genes, and a biologist may want to discover these through unsupervised model- ing of gene expression data. We present a new nonparametric Bayesian method, the In- finite Overlapping Mixture Model (IOMM), for modeling overlapping clusters. The IOMM uses exponential family distributions to model each cluster and forms an over- lapping mixture by taking products of such distributions, much like products of experts (Hinton, 2002). The IOMM allows an un- bounded number of clusters, and assignments of points to (multiple) clusters is modeled us- ing an Indian Buet Process (IBP), (Griths and Ghahramani, 2006). The IOMM has the desirable properties of being able to focus in on overlapping regions while maintaining the ability to model a potentially infinite num- ber of clusters which may overlap. We derive MCMC inference algorithms for the IOMM and show that these can be used to cluster movies into multiple genres.

Using Manifold Stucture for Partially Labeled Classification Abstract: We consider the general problem of utilizing both labeled and unlabeleddata to improve classification accuracy. Under the assumptionthat the data lie on a submanifold in a high dimensional space,we develop an algorithmic framework to classify a partially labeleddata set in a principled manner. The central idea of our approach isthat classification functions are naturally defined only on the submanifoldin question rather than the total ambient space. Using theLaplace Beltrami...

Modeling Human Motion Using Binary Latent Variables We propose a non-linear generative model for human motion data that uses an undirected model with binary latent variables and real-valued "visible" variables that represent joint angles. The latent and visible variabl es at each time step re- ceive directed connections from the visible variables at th e last few time-steps. Such an architecture makes on-line inference efficient and a llows us to use a sim- ple approximate learning procedure. After training, the model finds a single set of parameters that simultaneously capture several differe nt kinds of motion. We demonstrate the power of our approach by synthesizing various motion sequences and by performing on-line filling in of data lost during motio n capture. Website: http://www.cs.toronto.edu/∼gwtaylor/publications/nips2006mhmublv/

Dependent Fluents We discuss the persistence of the indirect ef­ fects of an action—the question when such ef­ fects are subject to the commonsense law of in­ ertia, and how to describe their evolution in the cases when inertia does not apply. Our model of nonpersistent effects involves the assumption that the value of the fluent in question is deter­ mined by the values of other fluents, although the dependency may be partially or completely unknown. This view leads us to a new high- level action language ARD (for Actions, Ram­ ifications and Dependencies) that is capable of describing both persistent and nonpersistent ef­ fects. Unlike the action languages introduced in the past, ARD is "non-Markovia n," in the sense that the evolution of the fluents described in this language may depend on their history, and not only on their current values.

The Tractable Cognition Thesis. The recognition that human minds/brains are finite systems with limited resources for computation has led some researchers to advance the Tractable Cognition thesis: Human cognitive capacities are constrained by computational tractability. This thesis, if true, serves cognitive psychology by constraining the space of computational-level theories of cognition. To utilize this constraint, a precise and workable definition of "computational tractability" is needed. Following computer science tradition, many cognitive scientists and psychologists define computational tractability as polynomial-time computability, leading to the P-Cognition thesis. This article explains how and why the P-Cognition thesis may be overly restrictive, risking the exclusion of veridical computational-level theories from scientific investigation. An argument is made to replace the P-Cognition thesis by the FPT-Cognition thesis as an alternative formalization of the Tractable Cognition thesis (here, FPT stands for fixed-parameter tractable). Possible objections to the Tractable Cognition thesis, and its proposed formalization, are discussed, and existing misconceptions are clarified.

Database storage management with object-based storage devices

Exploiting Web Log Mining for Web Cache Enhancement Improving the performance of the Web is a crucial requirement, since its popularity resulted in a large increase in the user perceived latency. In this paper, we describe a Web caching scheme that capitalizes on prefetching. Prefetching refers to the mechanism of deducing forthcoming page accesses of a client, based on access log information. Web log mining methods are exploited to provide effective prediction of Web-user accesses. The proposed scheme achieves a coordination between the two techniques (i.e., caching and prefetching). The prefetched documents are accommodated in a dedicated part of the cache, to avoid the drawback of incorrect replacement of requested documents. The requirements of the Web are taken into account, compared to the existing schemes for buffer management in database and operating systems. Experimental results indicate the superiority of the proposed method compared to the previous ones, in terms of improvement in cache performance.

Improving Citation Polarity Classification With Product Reviews Recent work classifying citations in scientific literature has shown that it is possible to improve classification results with extensive feature engineering. While this result confirms that citation classification is feasible, there are two drawbacks to this approach: (i) it requires a large annotated corpus for supervised classification, which in the case of scientific literature is quite expensive; and (ii) feature engineering that is too specific to one area of scientific literature may not be portable to other domains, even within scientific literature. In this paper we address these two drawbacks. First, we frame citation classification as a domain adaptation task and leverage the abundant labeled data available in other domains. Then, to avoid over-engineering specific citation features for a particular scientific domain, we explore a deep learning neural network approach that has shown to generalize well across domains using unigram and bigram features. We achieve better citation classification results with this cross-domain approach than using in-domain classification.

Soundness and completeness theorems for three formalizations of action Instead of trying to compare methodologies for reasoning about action on the basis of specific examples, we focus here on a general class of problems, expressible in a declarative language A. We propose three translations, P, R and B from A, representing respectively the first order methods of reasoning about action proposed by Pednault and Reiter and the circumscriptive approach of Baker. We then prove the soundness and completeness of these translations relative to the semantics of A. This lets us compare these three methods in a mathematically precise fashion. Moreover, we apply the methods of Baker in a general setting and prove a theorem which shows that if the domain of interest can be expressed in A, circumscription yields results which are intuitively expected.

Event calculus and temporal action logics compared We compare the event calculus and temporal action logics (TAL), two formalisms for reasoning about action and change. We prove that, if the formalisms are restricted to integer time, inertial fluents, and relational fluents, and if TAL action type specifications are restricted to definite reassignment of a single fluent, then the formalisms are not equivalent. We argue that equivalence cannot be restored by using more general TAL action type specifications. We prove however that, if the formalisms are further restricted to single-step actions, then they are logically equivalent.

A Realistic Experiment in Knowledge Representation in Open Event Calculus: Protocol Specification This paper presents one of the first realistic experiments in the use of Event Calculusin Open Logic Programming: the specification of a process protocol. The specification taskinvolves most of the common complications of temporal reasoning: the representation of contextdependent actions, of preconditions and ramifications of actions, the modelling of systemfaults, and most of all, the representation of uncertainty of actions. As the underlying language,the Open Logic Programming...

Knowledge Assimilation in Domains of Actions: A Possible Causes Approach . One major problem in the process of knowledge assimilation ishow to deal with inconsistency of new knowledge and the existing knowledgebase. In this paper we present a formal, provably correct and yet computationalmethodology for assimilation of new knowledge into knowledge basesabout actions and changes based on the slogan: What is believed is what isexplained. Technically, we employ Gelfond and Lifschitz" action description language A to describe domains of actions. The knowledge...

Filter preferential entailment for the logic of action in almost continuous worlds Mechanical systems, of the kinds which are of interest for qualitative reasoning, are characterized by a set of real-valued parameters, each of which is a piecewise continuous function of real-valued time. A temporal logic is introduced which allows the description of parameters, both in their continuous intervals and around their breakpoints, and which also allows the description of actions being performed in sequence or in parallel. If axioms are given which characterize physical laws, conditions and effects of actions, and observations or goals at specific points in time, one wishes to identify sets of actions ("plans") which account for the observations or obtain the goals. The paper proposes preference criteria which should determine the model set for such axioms. It is shown that conventional preferential entailment is not sufficient. A modified condition, filter preferential entailment is defined where preference conditions and axiom satisfaction conditions are interleaved.

Nonmonotonic reasoning in the framework of situation calculus Most of the solutions proposed to the Yale shooting problem haveeither introduced new nonmonotonic reasoning methods (generally involvingtemporal priorities) or completely reformulated the domainaxioms to represent causality explicitly. This paper presents a newsolution based on the idea that since the abnormality predicate takesa situational argument, it is important for the meanings of the situationsto be held constant across the various models being compared.This is accomplished by a...

An Efficient Unification Algorithm

Two components of an action language Some of the recent work on representing action makes use of high&dash;level action languages. In this paper we show that an action language can be represented as the sum of two distinct parts: an “action description language” and an “action query language.” A set of propositions in an action description language describes the effects of actions on states. Mathematically, it defines a transition system of the kind familiar from the theory of finite automata. An action query language serves for expressing properties of paths in a given transition system. We define the general concepts of a transition system, of an action description language and of an action query language, give a series of examples of languages of both kinds, and show how to combine a description language and a query language into one. This construction makes it possible to design the two components of an action language independently, which leads to the simplification and clarification of the theory of actions.

Conformant planning via symbolic model checking We tackle the problem of planning in nondeterministic domains, by presenting a new approach to conformant planning. Conformant planning is the problem of finding a sequence of actions that is guaranteed to achieve the goal despite the nondeterminism of the domain. Our approach is based on the representation of the planning domain as a finite state automaton. We use Symbolic Model Checking techniques, in particular Binary Decision Diagrams, to compactly represent and efficiently search the automaton. In this paper we make the following contributions. First, we present a general planning algorithm for conformant planning, which applies to fully nondeterministic domains, with uncertainty in the initial condition and in action effects. The algorithm is based on a breadth-first, backward search, and returns conformant plans of minimal length, if a solution to the planning problem exists, otherwise it terminates concluding that the problem admits no conformant solution. Second, we provide a symbolic representation of the search space based on Binary Decision Diagrams (BDDs), which is the basis for search techniques derived from symbolic model checking. The symbolic representation makes it possible to analyze potentially large sets of states and transitions in a single computation step, thus providing for an efficient implementation. Third, we present CMBP (Conformant Model Based Planner), an efficient implementation of the data structures and algorithm described above, directly based on BDD manipulations, which allows for a compact representation of the search layers and an efficient implementation of the search steps. Finally, we present an experimental comparison of our approach with the state-of-the-art conformant planners CGP, QBFPLAN and GPT. Our analysis includes all the planning problems from the distribution packages of these systems, plus other problems defined to stress a number of specific factors. Our approach appears to be the most effective: CMBP is strictly more expressive than QBFPLAN and CGP and, in all the problems where a comparison is possible, CMBP outperforms its competitors, sometimes by orders of magnitude.

An Abductive Proof Procedure for Reasoning About Actions in Modal Logic Programming In this paper we propose a modal approach for reasoning about actions in a logic programming framework. We introduce a modal language which makes use of abductive assumptions to deal with persistency, and provides a solution to the ramification problem, by allowing one-way “causal rules” to be defined among fluents.

Act Local, Think Global: Width Notions for Tractable Planning Many of the benchmark domains in AI planning are tractable on an individual basis. In this paper, we seek a theoretical, domain-independent explanation for their tractability. We present a family of structural conditions that both imply tractability and capture some of the es- tablished benchmark domains. These structural condi- tions are, roughly speaking, based on measures of how many variables need to be changed in order to move a state closer to a goal state.

Human-level control through deep reinforcement learning. The theory of reinforcement learning provides a normative account', deeply rooted in psychological' and neuroscientifie perspectives on animal behaviour, of how agents may optimize their control of an environment. To use reinforcement learning successfully in situations approaching real-world complexity, however, agents are confronted with a difficult task: they must derive efficient representations of the environment from high-dimensional sensory inputs, and use these to generalize past experience to new situations. Remarkably, humans and other animals seem to solve this problem through a harmonious combination of reinforcement learning and hierarchical sensory processing systems4'5, the former evidenced by a wealth of neural data revealing notable parallels between the phasic signals emitted by dopaminergic neurons and temporal difference reinforcement learning algorithms'. While reinforcement learning agents have achieved some successes in a variety of domains", their applicability has previously been limited to domains in which useful features can be handcrafted, or to domains with fully observed, low-dimensional state spaces. Here we use recent advances in training deep neural networks'" to develop a novel artificial agent, termed a deep Q-network, that can learn successful policies directly from high-dimensional sensory inputs using end-to-end reinforcement learning. We tested this agent on the challenging domain of classic Atari 2600 games". We demonstrate that the deep Q-network agent, receiving only the pixels and the game score as inputs, was able to surpass the performance of all previous algorithms and achieve a level comparable to that of a professional human games tester across a set of 49 games, using the same algorithm, network architecture and hyperparameters. This work bridges the divide between high-dimensional sensory inputs and actions, resulting in the first artificial agent that is capable of learning to excel at a diverse array of challenging tasks.

An Improved Long-Term File Usage Prediction Algorithm

Unsupervised (Parameter) Learning For Mrfs On Bipartite Graphs We consider unsupervised (parameter) learning for general Markov random fields on bipartite graphs. This model class includes Restricted Boltzmann Machines. We show that besides the widely used stochastic gradient approximation (a.k.a. Persistent Contrastive Divergence) there is an alternative learning approach - a modified EM algorithm which is tractable because of the bipartiteness of the model graph. We compare the resulting double loop algorithm and the PCD learning experimentally and show that the former converges faster and more stable than the latter.

Optimal Planning in the Presence of Conditional Effects: Extending LM-Cut with Context Splitting. The LM-Cut heuristic is currently the most successful heuristic in optimal STRIPS planning but it cannot be applied in the presence of conditional effects. Keyder, Hoffmann and Haslum recently showed that the obvious extensions to such effects ruin the nice theoretical properties of LM-Cut. We propose a new method based on context splitting that preserves these properties.

On the Feasibility of Planning Graph Style Heuristics for HTN Planning. In classical planning, the polynomial-time computability of propositional delete-free planning (planning with only positive effects and preconditions) led to the highly successful Relaxed Graphplan heuristic. We present a hierarchy of new computational complexity results for different classes of propositional delete-free HTN planning, with two main results: We prove that finding a plan for the delete-relaxation of a propositional HTN problem is NP-complete: hence unless P=NP, there is no directly analogous GraphPlan heuristic for HTN planning. However, a further relaxation of HTN planning (delete-free HTN planning with task insertion) is polynomial-time computable. Thus, there may be a possibility of using this or other relaxations to develop search heuristics for HTN planning.

Tight Bounds for HTN Planning. Although HTN planning is in general undecidable, there are many syntactically identifiable sub-classes of HTN problems that can be decided. For these sub-classes, the decision procedures provide upper complexity bounds. Lower bounds were often not investigated in more detail, however. We generalize a propositional HTN formalization to one that is based upon a function-free first-order logic and provide tight upper and lower complexity results along three axes: whether variables are allowed in operator and method schemas, whether the initial task and methods must be totally ordered, and where recursion is allowed (arbitrary recursion, tail-recursion, and acyclic problems). Our findings have practical implications, both for the reuse of classical planning techniques for HTN planning, and for the design of efficient HTN algorithms.

Trees of shortest paths vs. Steiner trees: understanding and improving delete relaxation heuristics Heuristic search using heuristics extracted from the delete relaxation is one of the most effective methods in planning. Since finding the optimal solution of the delete relaxation is intractable, various heuristics introduce independence assumptions, the implications of which are not yet fully understood. Here we use concepts from graph theory to show that in problems with unary action preconditions, the delete relaxation is closely related to the Steiner Tree problem, and that the independence assumption for the set of goals results in a tree-of-shortest-paths approximation. We analyze the limitations of this approximation and develop an alternative method for computing relaxed plans that addresses them. The method is used to guide a greedy best-first search, where it is shown to improve plan quality and coverage over several benchmark domains.

Ordered landmarks in planning Many known planning tasks have inherent constraints concerning the best order in which to achieve the goals. A number of research efiorts have been made to detect such constraints and to use them for guiding search, in the hope of speeding up the planning process. We go beyond the previous approaches by considering ordering constraints not only over the (top-level) goals, but also over the sub-goals that will necessarily arise during planning. Landmarks are facts that must be true at some point in every valid solution plan. We extend Koehler and Hoffmann's definition of reasonable orders between top level goals to the more general case of landmarks. We show how landmarks can be found, how their reasonable orders can be approximated, and how this information can be used to decompose a given planning task into several smaller sub-tasks. Our methodology is completely domain- and planner-independent. The implementation demonstrates that the approach can yield significant runtime performance improvements when used as a control loop around state-of-the-art sub-optimal planning systems, as exemplified by FF and LPG.

Planning as satisfiability: Heuristics Reduction to SAT is a very successful approach to solving hard combinatorial problems in Artificial Intelligence and computer science in general. Most commonly, problem instances reduced to SAT are solved with a general-purpose SAT solver. Although there is the obvious possibility of improving the SAT solving process with application-specific heuristics, this has rarely been done successfully. In this work we propose a planning-specific variable selection strategy for SAT solving. The strategy is based on generic principles about properties of plans, and its performance with standard planning benchmarks often substantially improves on generic variable selection heuristics, such as VSIDS, and often lifts it to the same level with other search methods such as explicit state-space search with heuristic search algorithms.

On the complexity of planning for agent teams and its implications for single agent planning If the complexity of planning for a single agent is described by some function f of the input, how much more difficult is it to plan for a team of n cooperating agents? If these agents are completely independent, we can simply solve n single agent problems, scaling linearly with the number of agents. But if all the agents interact tightly, we really need to solve a single problem that is n times larger, which could be exponentially (in n) harder to solve. Is a more general characterization possible? To formulate this question precisely, we minimally extend the standard STRIPS model to describe multi-agent planning problems. Then, we identify two problem parameters that help us answer our question. The first parameter is independent of the precise task the multi-agent system should plan for, and it captures the structure of the possible direct interactions between the agents via the tree-width of a graph induced by the team. The second parameter is task-dependent, and it captures the minimal number of interactions by the ''most interacting'' agent in the team that is needed to solve the problem. We show that multi-agent planning problems can be solved in time exponential only in these parameters. Thus, when these parameters are bounded, the complexity scales only polynomially in the size of the agent team. These results also have direct implications for the single-agent case: by casting single-agent planning tasks as multi-agent planning tasks, we can devise novel methods for decomposition-based planning for single agents. We analyze one such method, and use the techniques developed to provide some of the strongest tractability results for classical single-agent planning to date.

Utilizing Problem Structure in Planning: A Local Search Approach

Constructing conditional plans by a theorem-prover The research on conditional planning rejects the assumptions that there is no uncertainty or incompleteness of knowledge with respect to the state and changes of the system the plans operate on. Without these assumptions the sequences of operations that achieve the goals depend on the initial state and the outcomes of nondeterministic changes in the system. This setting raises the questions of how to represent the plans and how to perform plan search. The answers are quite different from those in the simpler classical framework. In this paper, we approach conditional planning from a new viewpoint that is motivated by the use of satisfiability algorithms in classical planning. Translating conditional planning to formulae in the propositional logic is not feasible because of inherent computational limitations. Instead, we translate conditional planning to quantified Boolean formulae. We discuss three formalizations of conditional planning as quantified Boolean formulae, and present experimental results obtained with a theorem-prover.

Filter preferential entailment for the logic of action in almost continuous worlds Mechanical systems, of the kinds which are of interest for qualitative reasoning, are characterized by a set of real-valued parameters, each of which is a piecewise continuous function of real-valued time. A temporal logic is introduced which allows the description of parameters, both in their continuous intervals and around their breakpoints, and which also allows the description of actions being performed in sequence or in parallel. If axioms are given which characterize physical laws, conditions and effects of actions, and observations or goals at specific points in time, one wishes to identify sets of actions ("plans") which account for the observations or obtain the goals. The paper proposes preference criteria which should determine the model set for such axioms. It is shown that conventional preferential entailment is not sufficient. A modified condition, filter preferential entailment is defined where preference conditions and axiom satisfaction conditions are interleaved.

The ellipsoid method and its consequences in combinatorial optimization. L. G. Khachiyan recently published a polynomial algorithm to check feasibility of a system of linear inequalities. The method is an adaptation of an algorithm proposed by Shor for non-linear optimization problems. In this paper we show that the method also yields interesting results in combinatorial optimization. Thus it yields polynomial algorithms for vertex packing in perfect graphs; for the matching and matroid intersection problems; for optimum covering of directed cuts of a digraph; for the minimum value of a submodular set function; and for other important combinatorial problems. On the negative side, it yields a proof that weighted fractional chromatic number is NP-hard.

Portable run-time support for dynamic object-oriented parallel processing Mentat is an object-oriented parallel processing system designed to simplify the task of writing portable parallel programs for parallel machines and workstation networks. The Mentat compiler and run-time system work together to automatically manage the communication and synchronization between objects. The run-time system marshals member function arguments, schedules objects on processors, and dynamically constructs and executes large-grain data dependence graphs. In this article we present the Mentat run-time system. We focus on three aspects—the software architecture, including the interface to the compiler and the structure and interaction of the principle components of the run-time system; the run-time overhead on a component-by-component basis for two platforms, a Sun SparcStation 2 and an Intel Paragon; and an analysis of the minimum granularity required for application programs to overcome the run-time overhead.

Building extensible frameworks for data processing: The case of MDP, Modular toolkit for Data Processing. Data processing is a ubiquitous task in scientific research, and much energy is spent on the development of appropriate algorithms. It is thus relatively easy to find software implementations of the most common methods. On the other hand, when building concrete applications, developers are often confronted with several additional chores that need to be carried out beside the individual processing steps. These include for example training and executing a sequence of several algorithms, writing code that can be executed in parallel on several processors, or producing a visual description of the application. The Modular toolkit for Data Processing (MDP) is an open source Python library that provides an implementation of several widespread algorithms and offers a unified framework to combine them to build more complex data processing architectures. In this paper we concentrate on some of the newer features of MOP, focusing on the choices made to automatize repetitive tasks for users and developers. In particular, we describe the support for parallel computing and how this is implemented via a flexible extension mechanism. We also briefly discuss the support for algorithms that require bi-directional data flow. (C) 2011 Elsevier B.V. All rights reserved.

Learning A Lexical Simplifier Using Wikipedia In this paper we introduce a new lexical simplification approach. We extract over 30K candidate lexical simplifications by identifying aligned words in a sentence-aligned corpus of English Wikipedia with Simple English Wikipedia. To apply these rules, we learn a feature-based ranker using SVMnk trained on a set of labeled simplifications collected using Amazon's Mechanical Turk. Using human simplifications for evaluation, we achieve a precision of 76% with changes in 86% of the examples.

Learning Generalized Policies from Planning Examples Using Concept Languages In this paper we are concerned with the problem of learning how to solve planning problems in one domain given a number of solved instances. This problem is formulated as the problem of inferring a function that operates over all instances in the domain and maps states and goals into actions. We call such functions generalized policies and the question that we address is how to learn suitable representations of generalized policies from data. This question has been addressed recently by Roni Khardon (Technical Report TR-09-97, Harvard, 1997). Khardon represents generalized policies using an ordered list of existentially quantified rules that are inferred from a training set using a version of Rivest's learning algorithm (Machine Learning, vol. 2, no. 3, pp. 229–246, 1987). Here, we follow Khardon's approach but represent generalized policies in a different way using a concept language. We show through a number of experiments in the blocks-world that the concept language yields a better policy using a smaller set of examples and no background knowledge.

Learning action strategies for planning domains This paper reports on experiments where techniques of supervised machine learning areapplied to the problem of planning. The input to the learning algorithm is composedof a description of a planning domain, planning problems in this domain, and solutionsfor them. The output is an efficient algorithm --- a strategy --- for solving problems inthat domain. We test the strategy on an independent set of planning problems fromthe same domain, so that success is measured by its ability to solve...

Constituent Parsing with Incremental Sigmoid Belief Networks We introduce a framework for syntactic parsing with latent variables based on a form of dynamic Sigmoid Belief Networks called Incremental Sigmoid Belief Networks. We demonstrate that a previous feed-forward neural network parsing model can be viewed as a coarse approximation to inference with this class of graphical model. By construct- ing a more accurate but still tractable ap- proximation, we significantly improve pars- ing accuracy, suggesting that ISBNs provide a good idealization for parsing. This gener- ative model of parsing achieves state-of-the- art results on WSJ text and 8% error reduc- tion over the baseline neural network parser.

From Conformant into Classical Planning: Efficient Translations that May Be Complete Too Focusing on the computation of conformant plans whose verification can be done efficiently, we have recently proposed a polynomial scheme for mapping conformant problems P with deterministic actions into classical problems K(P). The scheme is sound as the classical plans are all conformant, but is incomplete as the converse relation does not always hold. In this paper, we extend this work and consider an alternative, more powerful translation based on the introduction of epistemic tagged literals KL/t where L is a literal in P and t is a set of literals in P unknown in the initial situation. The translation ensures that a plan makes KL/t true only when the plan makes L certain in P given the assumption that t is initially true. We show that under general conditions the new translation scheme is complete and that its complexity can be characterized in terms of a parameter of the problem that we call conformant width. We show that the complexity of the translation is exponential in the problem width only, find that the width of almost all benchmarks is 1, and show that a conformant planner based on this translation solves some interesting domains that cannot be solved by other planners. This translation is the basis for T_0, the best performing planner in the Conformant Track of the 2006 International Planning Competition.

Conformant Graphplan Planning under uncertainty is a difficult task. If sensory information is available, it is possible to do contingency planning - that is, develop plans where certain branches are executed conditionally, based on the outcome of sensory actions. However, even without sensory information, it is often possible to develop useful plans that succeed no matter which of the allowed states the world is actually in. We refer to this type of planning as conformant planning.Few conformant planners have been built, partly because conformant planning requires the ability to reason about disjunction. In this paper we describe Conformant Graphplan (CGP), a Graphplan-based planner that develops sound (non-contingent) plans when faced with uncertainty in the initial conditions and in the outcome of actions. The basic idea is to develop separate plan graphs for each possible world. This requires some subtle changes to both the graph expansion and solution extraction phases of Oraphplan. In particular, the solution extraction phase must consider the unexpected side effects of actions in other possible worlds, and must confront any undesirable effects.We show that COP performs significantly better than two previous (probabilistic) conformant planners.

Backpropagation Applied to Handwritten Zip Code Recognition. The ability of learning networks to generalize can be greatly enhanced by providing constraints from the task domain. This paper demonstrates how such constraints can be integrated into a backpropagation network through the architecture of the network. This approach has been successfully applied to the recognition of handwritten zip code digits provided by the U.S. Postal Service. A single network learns the entire recognition operation, going from the normalized image of the character to the final classification.

Feasibility of a serverless distributed file system deployed on an existing set of desktop PCs We consider an architecture for a serverless distributed file system that does not assume mutual trust among the client computers. The system provides security, availability, and reliability by distributing multiple encrypted replicas of each file among the client machines. To assess the feasibility of deploying this system on an existing desktop infrastructure, we measure and analyze a large set of client machines in a commercial environment. In particular, we measure and report results on disk usage and content; file activity; and machine uptimes, lifetimes, and loads. We conclude that the measured desktop infrastructure would passably support our proposed system, providing availability on the order of one unfilled file request per user per thousand days.

The HP AutoRAID hierarchical storage system Configuring redundant disk arrays is a black art. To configure an array properly, a system administrator must understand the details of both the array and the workload it will support. Incorrect understanding of either, or changes in the workload over time, can lead to poor performance. We present a solution to this problem: a two-level storage hierarchy implemented inside a single disk-array controller. In the upper level of this hierarchy, two copies of active data are stored to provide full redundancy and excellent performance. In the lower level, RAID 5 parity protection is used to provide excellent storage cost for inactive data, at somewhat lower performance. The technology we describe in this article, know as HP AutoRAID, automatically and transparently manages migration of data blocks between these two levels as access patterns change. The result is a fully redundant storage system that is extremely easy to use, is suitable for a wide variety of workloads, is largely insensitive to dynamic workload changes, and performs much better than disk arrays with comparable numbers of spindles and much larger amounts of front-end RAM cache. Because the implementation of the HP AutoRAID technology is almost entirely in software, the additional hardware cost for these benefits is very small. We describe the HP AutoRAID technology in detail, provide performance data for an embodiment of it in a storage array, and summarize the results of simulation studies used to choose algorithms implemented in the array.

The Complexity of Global Constraints We study the computational complexity of reasoning with global constraints. We show that reasoning with such constraints is intractable in general. We then demonstrate how the same tools of computational com- plexity can be used in the design and analysis of spe- cific global constraints. In particular, we illustrate how computational complexity can be used to determine when a lesser level of local consistency should be en- forced, when decomposing constraints will lose prun- ing, and when combining constraints is tractable. We also show how the same tools can be used to study symmetry breaking, meta-constraints like the cardinal- ity constraint, and learning nogoods.

Serverless network file systems We propose a new paradigm for network file system design: serverless network file systems. While traditional network file systems rely on a central server machine, a serverless system utilizes workstations cooperating as peers to provide all file system services. Any machine in the system can store, cache, or control any block of data. Our approach uses this location independence, in combination with fast local area networks, to provide better performance and scalability than traditional file systems. Furthermore, because any machine in the system can assume the responsibilities of a failed component, our serverless design also provides high availability via redundatn data storage. To demonstrate our approach, we have implemented a prototype serverless network file system called xFS. Preliminary performance measurements suggest that our architecture achieves its goal of scalability. For instance, in a 32-node xFS system with 32 active clients, each client receives nearly as much read or write throughput as it would see if it were the only active client.

Simultaneous Pipelining in QPipe: Exploiting Work Sharing Opportunities Across Queries Data warehousing and scientific database applications operate on massive datasets and are characterized by complex queries accessing large portions of the database. Concurrent queries often exhibit high data and computation overlap, e.g., they access the same relations on disk, compute similar aggregates, or share intermediate results. Unfortunately, run-time sharing in modern database engines is limited by the paradigm of invoking an independent set of operator instances per query, potentially missing sharing opportunities if the buffer pool evicts data early.

The Performance Impact of Kernel Prefetching on Buffer Cache Replacement Algorithms A fundamental challenge in improving file system performance is to design effective block replacement algorithms to minimize buffer cache misses. Despite the well-known interactions between prefetching and caching, almost all buffer cache replacement algorithms have been proposed and studied comparatively, without taking into account file system prefetching, which exists in all modern operating systems. This paper shows that such kernel prefetching can have a significant impact on the relative performance in terms of the number of actual disk I/Os of many well-known replacement algorithms; it can not only narrow the performance gap but also change the relative performance benefits of different algorithms. Moreover, since prefetching can increase the number of blocks clustered for each disk I/O and, hence, the time to complete the I/O, the reduction in the number of disk I/Os may not translate into proportional reduction in the total I/O time. These results demonstrate the importance of buffer caching research taking file system prefetching into consideration and comparing the actual disk I/Os and the execution time under different replacement algorithms.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

The Case for Cross-Layer Optimizations in Storage: A Workflow-Optimized Storage System This paper proposes using file system custom metadata as a bidirectional communication channel between applications and the storage system. This channel can be used to pass hints that enable cross-layer optimizations, an option hindered today by the ossified file-system interface. We study this approach in context of storage system support for large-scale workflow execution systems: Our workflow optimized storage system (WOSS), exploits application hints to provide per-file optimized operations, and exposes data location to enable location-aware scheduling. This paper argues that an incremental adoption path for adopting cross-layer optimizations in storage systems exists, presents the system architecture for a workflow-optimized storage system and its integration with a workflow runtime engine, and evaluates the proposed approach using synthetic as well as real applications workloads.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Flash caching on the storage client Flash memory has recently become popular as a caching medium. Most uses to date are on the storage server side. We investigate a different structure: flash as a cache on the client side of a networked storage environment. We use trace-driven simulation to explore the design space. We consider a wide range of configurations and policies to determine the potential client-side caches might offer and how best to arrange them. Our results show that the flash cache writeback policy does not significantly affect performance. Write-through is sufficient; this greatly simplifies cache consistency handling. We also find that the chief benefit of the flash cache is its size, not its persistence. Cache persistence offers additional performance benefits at system restart at essentially no runtime cost. Finally, for some workloads a large flash cache allows using miniscule amounts of RAM for file caching (e.g., 256 KB) leaving more memory available for application use.

Janus: optimal flash provisioning for cloud storage workloads Janus is a system for partitioning the flash storage tier between workloads in a cloud-scale distributed file system with two tiers, flash storage and disk. The file system stores newly created files in the flash tier and moves them to the disk tier using either a First-In-First-Out (FIFO) policy or a Least-Recently-Used (LRU) policy, subject to per-workload allocations. Janus constructs compact metrics of the cacheability of the different workloads, using sampled distributed traces because of the large scale of the system. From these metrics, we formulate and solve an optimization problem to determine the flash allocation to workloads that maximizes the total reads sent to the flash tier, subject to operator-set priorities and bounds on flash write rates. Using measurements from production workloads in multiple data centers using these recommendations, as well as traces of other production workloads, we show that the resulting allocation improves the flash hit rate by 47-76% compared to a unified tier shared by all workloads. Based on these results and an analysis of several thousand production workloads, we conclude that flash storage is a cost-effective complement to disks in data centers.

PCM-Based Durable Write Cache for Fast Disk I/O Flash based solid-state devices (FSSDs) have been adopted within the memory hierarchy to improve the performance of hard disk drive (HDD) based storage system. However, with the fast development of storage-class memories, new storage technologies with better performance and higher write endurance than FSSDs are emerging, e.g., phase-change memory (PCM). Understanding how to leverage these state-of the-art storage technologies for modern computing systems is important to solve challenging data intensive computing problems. In this paper, we propose to leverage PCM for a hybrid PCM-HDD storage architecture. We identify the limitations of traditional LRU caching algorithms for PCMbased caches, and develop a novel hash-based write caching scheme called HALO to improve random write performance of hard disks. To address the limited durability of PCM devices and solve the degraded spatial locality in traditional wear-leveling techniques, we further propose novel PCM management algorithms that provide effective wear-leveling while maximizing access parallelism. We have evaluated this PCM-based hybrid storage architecture using applications with a diverse set of I/O access patterns. Our experimental results demonstrate that the HALO caching scheme leads to an average reduction of 36.8% in execution time compared to the LRU caching scheme, and that the SFC wear leveling extends the lifetime of PCM by a factor of 21.6.

Caching less for better performance: balancing cache size and update cost of flash memory cache in hybrid storage systems Hybrid storage solutions use NAND flash memory based Solid State Drives (SSDs) as non-volatile cache and traditional Hard Disk Drives (HDDs) as lower level storage. Unlike a typical cache, internally, the flash memory cache is divided into cache space and overprovisioned space, used for garbage collection. We show that balancing the two spaces appropriately helps improve the performance of hybrid storage systems. We show that contrary to expectations, the cache need not be filled with data to the fullest, but may be better served by reserving space for garbage collection. For this balancing act, we present a dynamic scheme that further divides the cache space into read and write caches and manages the three spaces according to the workload characteristics for optimal performance. Experimental results show that our dynamic scheme improves performance of hybrid storage solutions up to the off-line optimal performance of a fixed partitioning scheme. Furthermore, as our scheme makes efficient use of the flash memory cache, it reduces the number of erase operations thereby extending the lifetime of SSDs.

Evaluation techniques for storage hierarchies The design of efficient storage hierarchies generally involves the repeated running of "typical" program address traces through a simulated storage system while various hierarchy design parameters are adjusted. This paper describes a new and efficient method of determining, in one pass of an address trace, performance measures for a large class of demand-paged, multilevel storage systems utilizing a variety of mapping schemes and replacement algorithms. The technique depends on an algorithm classification, called "stack algorithms," examples of which are "least frequently used," "least recently used," "optimal," and "random replacement" algorithms. The techniques yield the exact access frequency to each storage device, which can be used to estimate the overall performance of actual storage hierarchies.

Storage-Aware Caching: Revisiting Caching for Heterogeneous Storage Systems Modern storage environments are composed of a variety of devices with different performance characteristics. In this paper we explore storage-aware caching algorithms, in which the file buffer replacement algorithm explicitly accounts for differences in performance across devices. We introduce a new family of storage-aware caching algorithms that partition the cache, with one partition per device. The algorithms set the partition sizes dynamically to balance work across the devices. Through simulation, we show that our storage-aware policies perform similarly to LANDLORD, a cost-aware algorithm previously shown to perform well in Web caching environments. We also demonstrate that partitions can be easily incorporated into the Clock replacement algorithm, thus increasing the likelihood of deploying cost-aware algorithms in modern operating systems.

Informed prefetching and caching The underutilization of disk parallelism and file cache buffers by traditional file systems induces I/O stall time that degrades the performance of modern microprocessor-based systems. In this paper, we present aggressive mechanisms that tailor file system resource management to the needs of I/O-intensive applications. In particular, we show how to use application-disclosed access patterns (hints) to expose and exploit I/O parallelism and to allocate dynamically file buffers among three competing demands: prefetching hinted blocks, caching hinted blocks for reuse, and caching recently used data for unhinted accesses. Our approach estimates the impact of alternative buffer allocations on application execution time and applies a cost-benefit analysis to allocate buffers where they will have the greatest impact. We implemented informed prefetching and caching in DEC''s OSF/1 operating system and measured its performance on a 150 MHz Alpha equipped with 15 disks running a range of applications including text search, 3D scientific visualization, relational database queries, speech recognition, and computational chemistry. Informed prefetching reduces the execution time of the first four of these applications by 20% to 87%. Informed caching reduces the execution time of the fifth application by up to 30%.

On the scale and performance of cooperative Web proxy caching Abstract While algorithms for cooperative proxy caching have been widely studied, little is understood about cooperative- caching performance,in the large-scale World Wide Web en- vironment. This paper uses both trace-based analysis and analytic modelling,to show,the potential advantages and drawbacks of inter-proxy cooperation. With our traces, we evaluate quantitatively the performance-improvement po- tential of cooperation between 200 small-organization prox- ies within a university environment, and between two large- organization proxies handling 23,000 and 60,000 clients, re- spectively. With our model, we extend beyond these popula- tions to project cooperative caching behavior in regions with millions of clients. Overall, we demonstrate that cooperative caching has performance,benefits only within limited popu- lation bounds. We also use our model to examine the impli- cations of future trends in Web-access behavior and traffic.

Adaptive block rearrangement An adaptive technique for reducing disk seek times is described. The technique copies frequently referenced blocks from their original locations to reserved space near the middle of the disk. Reference frequencies need not be known in advance. Instead, they are estimated by monitoring the stream of arriving requests. Trace-driven simulations show that seek times can be cut substantially by copying only a small number of blocks using this technique. The technique has been implemented by modifying a UNIX device driver. No modifications are required to the file system that uses the driver.

Pushing the envelope: planning, propositional logic, and stochastic search Planning is a notoriously hard combinatorial search problem. In many interesting domains, current planning algorithms fail to scale up gracefully. By combining a general, stochastic search algorithm and appropriate problem encodings based on propositional logic, we are able to solve hard planning problems many times faster than the best current planning systems. Although stochastic methods have been shown to be very effective on a wide range of scheduling problems, this is the first demonstration of its power on truly challenging classical planning instances. This work also provides a new perspective on representational issues in planning.

Why does file system prefetching work? Most file systems attempt to predict which disk blocks will be needed in the near future and prefetch them into memory; this technique can improve application throughput as much as 50%. But why? The reasons include that the disk cache comes into play, the device driver amortizes the fixed cost of an I/O operation over a larger amount of data, total disk seek time can be decreased, and that programs can overlap computation and I/O. However, intuition does not tell us the relative benefit of each of these causes, or techniques for increasing the effectiveness of prefetching. To answer these questions, we constructed an analytic performance model for file system reads. The model is based on a 4.4BSD-derived file system, and parameterized by the access patterns of the files, layout of files on disk, and the design characteristics of the file system and of the underlying disk. We then validated the model against several simple workloads; the predictions of our model were typically within 4% of measured values, and differed at most by 9% from measured values. Using the model and experiments, we explain why and when prefetching works, and make proposals for how to tune file system and disk parameters to improve overall system throughput.

File system aging—increasing the relevance of file system benchmarks Benchmarks are important because they provide a means for users and researchers to characterize how their workloads will perform on different systems and different system architectures. The field of file system design is no different from other areas of research in this regard, and a variety of file system benchmarks are in use, representing a wide range of the different user workloads that may be run on a file system. A realistic benchmark, however, is only one of the tools that is required in order to understand how a file system design will perform in the real world. The benchmark must also be executed on a realistic file system. While the simplest approach may be to measure the performance of an empty file system, this represents a state that is seldom encountered by real users. In order to study file systems in more representative conditions, we present a methodology for aging a test file system by replaying a workload similar to that experienced by a real file system over a period of many months, or even years. Our aging tools allow the same aging workload to be applied to multiple versions of the same file system, allowing scientific evaluation of the relative merits of competing file system designs.In addition to describing our aging tools, we demonstrate their use by applying them to evaluate two enhancements to the file layout policies of the UNIX fast file system.

Global Reinforcement Learning in Neural Networks with Stochastic Synapses We have found a more general formulation of the REINFORCE learning principle which had been proposed by R. J. Williams for the case of artificial neural networks with stochastic cells ("Boltzmann machines"). This formulation has enabled us to apply the principle to global reinforcement learning in networks with deterministic neural cells but stochastic synapses, and to suggest two groups of new learning rules for such networks, including simple local rules. Numerical simulations have shown that at least for several popular benchmark problems one of the new learning rules may provide results on a par with the best known global reinforcement techniques.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Deep Online Hierarchical Unsupervised Learning for Pattern Mining from Utility Usage Data. Machine learning approaches for non-intrusive load monitoring (NILM) have focused on supervised algorithms. Unsupervised approaches can be more interesting and of more practical use in real case scenarios. More specifically, they do not require labelled training data to be collected from individual appliances and the algorithm can be deployed to operate on the measured aggregate data directly. In this paper, we propose a fully unsupervised NILM framework based on Deep Belief network (DBN) and online Latent Dirichlet Allocation (LDA). Firstly, the raw signals of the house utilities are fed into DBN to extract low-level generic features in an unsupervised fashion, and then the hierarchical Bayesian model, LDA, learns high-level features that capture the correlations between the low-level ones. Thus, the proposed method (DBN-LDA) harnesses the DBN's ability of learning distributed hierarchies of features to extract sophisticated appliances-specific features without the need of precise human-crafted input representations. The clustering power of the hierarchical Bayesian models helps further summarise the input data by extracting higher-level information representing the residents' energy consumption patterns. Using Deep-Hierarchical models reduces the computational complexity since LDA is not directly applied to the raw data. The computational efficiency is crucial as our application involves massive data from different types of utility usages. Moreover, we develop a novel online inference algorithm to cope with this big data. Another novelty of this work is that the data is a combination of different utilities (e.g., electricity, water and gas) and some sensors measurements. Finally, we propose different methods to evaluate the results and preliminary experiments show that the DBN-LDA is promising to extract useful patterns.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Learning Topic Representation For Smt With Neural Networks Statistical Machine Translation (SMT) usually utilizes contextual information to disambiguate translation candidates. However, it is often limited to contexts within sentence boundaries, hence broader topical information cannot be leveraged. In this paper, we propose a novel approach to learning topic representation for parallel data using a neural network architecture, where abundant topical contexts are embedded via topic relevant monolingual data. By associating each translation rule with the topic representation, topic relevant rules are selected according to the distributional similarity with the source text during SMT decoding. Experimental results show that our method significantly improves translation accuracy in the NIST Chinese-to-English translation task compared to a state-of-the-art baseline.

Market impact analysis via deep learned architectures How to deeply process market data sources and build systems to process accurate market impact analysis is an attractive problem. In this paper, we build up a system that exploits deep learning architecture to improve feature representations, and adopt state-of-the-art supervised learning algorithm—extreme learning machine—to predict market impacts. We empirically evaluate the performance of the system by comparing different configurations of representation learning and classification algorithms, and conduct experiments on the intraday tick-by-tick price data and corresponding commercial news archives of stocks in Hong Kong Stock Exchange. From the results, we find that in order to make system achieve good performance, both the representation learning and the classification algorithm play important roles, and comparing with various benchmark configurations of the system, deep learned feature representation together with extreme learning machine can give the highest market impact prediction accuracy.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Learning Topic Representation For Smt With Neural Networks Statistical Machine Translation (SMT) usually utilizes contextual information to disambiguate translation candidates. However, it is often limited to contexts within sentence boundaries, hence broader topical information cannot be leveraged. In this paper, we propose a novel approach to learning topic representation for parallel data using a neural network architecture, where abundant topical contexts are embedded via topic relevant monolingual data. By associating each translation rule with the topic representation, topic relevant rules are selected according to the distributional similarity with the source text during SMT decoding. Experimental results show that our method significantly improves translation accuracy in the NIST Chinese-to-English translation task compared to a state-of-the-art baseline.

Classification Experiments of DNA Sequences by Using a Deep Neural Network and Chaos Game Representation. Analysis and classification of sequences is one of the key research areas in bioinformatics. The basic tool for sequence analysis is alignment, but there are also other techniques that can be used. Frequency Chaos Game Representation is a technique that builds an image characteristic of the sequence The paper describes the first experiment in the use of a deep neural network for classification of DNA sequences represented as images by using the Frequency Chaos Game Representation.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Learning Topic Representation For Smt With Neural Networks Statistical Machine Translation (SMT) usually utilizes contextual information to disambiguate translation candidates. However, it is often limited to contexts within sentence boundaries, hence broader topical information cannot be leveraged. In this paper, we propose a novel approach to learning topic representation for parallel data using a neural network architecture, where abundant topical contexts are embedded via topic relevant monolingual data. By associating each translation rule with the topic representation, topic relevant rules are selected according to the distributional similarity with the source text during SMT decoding. Experimental results show that our method significantly improves translation accuracy in the NIST Chinese-to-English translation task compared to a state-of-the-art baseline.

Deep learning for visual understanding: A review. Deep learning algorithms are a subset of the machine learning algorithms, which aim at discovering multiple levels of distributed representations. Recently, numerous deep learning algorithms have been proposed to solve traditional artificial intelligence problems. This work aims to review the state-of-the-art in deep learning algorithms in computer vision by highlighting the contributions and challenges from over 210 recent research papers. It first gives an overview of various deep learning approaches and their recent developments, and then briefly describes their applications in diverse vision tasks, such as image classification, object detection, image retrieval, semantic segmentation and human pose estimation. Finally, the paper summarizes the future trends and challenges in designing and training deep neural networks.

A survey of machine learning for big data processing. There is no doubt that big data are now rapidly expanding in all science and engineering domains. While the potential of these massive data is undoubtedly significant, fully making sense of them requires new ways of thinking and novel learning techniques to address the various challenges. In this paper, we present a literature survey of the latest advances in researches on machine learning for big data processing. First, we review the machine learning techniques and highlight some promising learning methods in recent studies, such as representation learning, deep learning, distributed and parallel learning, transfer learning, active learning, and kernel-based learning. Next, we focus on the analysis and discussions about the challenges and possible solutions of machine learning for big data. Following that, we investigate the close connections of machine learning with signal processing techniques for big data processing. Finally, we outline several open issues and research trends.

3D object understanding with 3D Convolutional Neural Networks Feature engineering plays an important role in object understanding. Expressive discriminative features can guarantee the success of object understanding tasks. With remarkable ability of data abstraction, deep hierarchy architecture has the potential to represent objects. For 3D objects with multiple views, the existing deep learning methods can not handle all the views with high quality. In this paper, we propose a 3D convolutional neural network, a deep hierarchy model which has a similar structure with convolutional neural network. We employ stochastic gradient descent (SGD) method to pretrain the convolutional layer, and then a back-propagation method is proposed to fine-tune the whole network. Finally, we use the result of the two phases for 3D object retrieval. The proposed method is shown to out-perform the state-of-the-art approaches by experiments conducted on publicly available 3D object datasets.

A review on deep learning for recommender systems: challenges and remedies Recommender systems are effective tools of information filtering that are prevalent due to increasing access to the Internet, personalization trends, and changing habits of computer users. Although existing recommender systems are successful in producing decent recommendations, they still suffer from challenges such as accuracy, scalability, and cold-start. In the last few years, deep learning, the state-of-the-art machine learning technique utilized in many complex tasks, has been employed in recommender systems to improve the quality of recommendations. In this study, we provide a comprehensive review of deep learning-based recommendation approaches to enlighten and guide newbie researchers interested in the subject. We analyze compiled studies within four dimensions which are deep learning models utilized in recommender systems, remedies for the challenges of recommender systems, awareness and prevalence over recommendation domains, and the purposive properties. We also provide a comprehensive quantitative assessment of publications in the field and conclude by discussing gained insights and possible future work on the subject.

Improving Content-based and Hybrid Music Recommendation using Deep Learning Existing content-based music recommendation systems typically employ a \\textit{two-stage} approach. They first extract traditional audio content features such as Mel-frequency cepstral coefficients and then predict user preferences. However, these traditional features, originally not created for music recommendation, cannot capture all relevant information in the audio and thus put a cap on recommendation performance. Using a novel model based on deep belief network and probabilistic graphical model, we unify the two stages into an automated process that simultaneously learns features from audio content and makes personalized recommendations. Compared with existing deep learning based models, our model outperforms them in both the warm-start and cold-start stages without relying on collaborative filtering (CF). We then present an efficient hybrid method to seamlessly integrate the automatically learnt features and CF. Our hybrid method not only significantly improves the performance of CF but also outperforms the traditional feature mbased hybrid method.

A Novel Rbf Training Algorithm For Short-Term Electric Load Forecasting And Comparative Studies Because of their excellent scheduling capabilities, artificial neural networks (ANNs) are becoming popular in short-term electric power system forecasting, which is essential for ensuring both efficient and reliable operations and full exploitation of electrical energy trading as well. For such a reason, this paper investigates the effectiveness of some of the newest designed algorithms in machine learning to train typical radial basis function (RBF) networks for 24-h electric load forecasting: support vector regression (SVR), extreme learning machines (ELMs), decay RBF neural networks (DRNNs), improves second order, and error correction, drawing some conclusions useful for practical implementations.

A restricted Boltzmann machine based two-lead electrocardiography classification An restricted Boltzmann machine learning algorithm were proposed in the two-lead heart beat classification problem. ECG classification is a complex pattern recognition problem. The unsupervised learning algorithm of restricted Boltzmann machine is ideal in mining the massive unlabelled ECG wave beats collected in the heart healthcare monitoring applications. A restricted Boltzmann machine (RBM) is a generative stochastic artificial neural network that can learn a probability distribution over its set of inputs. In this paper a deep belief network was constructed and the RBM based algorithm was used in the classification problem. Under the recommended twelve classes by the ANSI/AAMI EC57: 1998/(R)2008 standard as the waveform labels, the algorithm was evaluated on the two-lead ECG dataset of MIT-BIH and gets the performance with accuracy of 98.829%. The proposed algorithm performed well in the two-lead ECG classification problem, which could be generalized to multi-lead unsupervised ECG classification or detection problems.

Semantic hashing We show how to learn a deep graphical model of the word-count vectors obtained from a large set of documents. The values of the latent variables in the deepest layer are easy to infer and give a much better representation of each document than Latent Semantic Analysis. When the deepest layer is forced to use a small number of binary variables (e.g. 32), the graphical model performs ''semantic hashing'': Documents are mapped to memory addresses in such a way that semantically similar documents are located at nearby addresses. Documents similar to a query document can then be found by simply accessing all the addresses that differ by only a few bits from the address of the query document. This way of extending the efficiency of hash-coding to approximate matching is much faster than locality sensitive hashing, which is the fastest current method. By using semantic hashing to filter the documents given to TF-IDF, we achieve higher accuracy than applying TF-IDF to the entire document set.

Nonparametric belief propagation for self-localization of sensor networks Automatic self-localization is a critical need for the effective use of ad hoc sensor networks in military or civilian applications. In general, self-localization involves the combination of absolute location information (e.g., from a global positioning system) with relative calibration information (e.g., distance measurements between sensors) over regions of the network. Furthermore, it is generally desirable to distribute the computational burden across the network and minimize the amount of intersensor communication. We demonstrate that the information used for sensor localization is fundamentally local with regard to the network topology and use this observation to reformulate the problem within a graphical model framework. We then present and demonstrate the utility of nonparametric belief propagation (NBP), a recent generalization of particle filtering, for both estimating sensor locations and representing location uncertainties. NBP has the advantage that it is easily implemented in a distributed fashion, admits a wide variety of statistical models, and can represent multimodal uncertainty. Using simulations of small to moderately sized sensor networks, we show that NBP may be made robust to outlier measurement errors by a simple model augmentation, and that judicious message construction can result in better estimates. Furthermore, we provide an analysis of NBP's communications requirements, showing that typically only a few messages per sensor are required, and that even low bit-rate approximations of these messages can be used with little or no performance impact.

Learning with local and global consistency We consider the general problem of learning from labeled and unlabeled data, which is often called semi-supervised learning or transductive inference. A principled approach to semi-supervised learning is to design a classifying function which is sufficiently smooth with respect to the intrinsic structure collectively revealed by known labeled and unlabeled points. We present a simple algorithm to obtain such a smooth solution. Our method yields encouraging experimental results on a number of classification problems and demonstrates effective use of unlabeled data.

A Downward Translation in the Polynomial Hierarchy Downward collapse (a.k.a. upward separation) refers to cases where the equality of two larger classes implies the equality of two smaller classes. We provide an unqualified downward collapse result completely within the polynomial hierarchy. In particular, we prove that, for k

Narratives as Programs

S/390 CMOS server I/O: The continuing evolution IBM has developed a strategy to achieve the high I/O demands of large servers. In a new environment of industry-standard peripheral component interconnect (PCI) attached adapters conforming to open I/O interfaces, S/390® has developed an efficient method of quickly integrating disk storage, communications, and future adapters. Preserving the S/390 I/O programming model and the high level of data integrity expected in S/390 products and reducing development cycle time and resources have further constrained design options. At the same time, S/390 developers have redesigned the traditional I/O components into the latest chip technologies. The developers have also designed a new internal link (STI) to meet the increased I/O bandwidth and connectivity required by the high processor performance of the third and fourth generations of S/390 CMOS servers. This paper describes this strategy and how it has led to systems that retain the differentiating features of S/390 products.

GPU-accelerated exhaustive search for third-order epistatic interactions in case-control studies. Interest in discovering combinations of genetic markers from case-control studies, such as Genome Wide Association Studies (GWAS), that are strongly associated to diseases has increased in recent years. Detecting epistasis, i.e. interactions among k markers (k >= 2), is an important but time consuming operation since statistical computations have to be performed for each k-tuple of measured markers. Efficient exhaustive methods have been proposed for k = 2, but exhaustive third-order analyses are thought to be impractical due to the cubic number of triples to be computed. Thus, most previous approaches apply heuristics to accelerate the analysis by discarding certain triples in advance. Unfortunately, these tools can fail to detect interesting interactions. We present GPU3SNP, a fast GPU-accelerated tool to exhaustively search for interactions among all marker-triples of a given case-control dataset. Our tool is able to analyze an input dataset with tens of thousands of markers in reasonable time thanks to two efficient CUDA kernels and efficient workload distribution techniques. For instance, a dataset consisting of 50,000 markers measured from 1000 individuals can be analyzed in less than 22h on a single compute node with 4 NVIDIA GTX Titan boards. (C) 2015 Elsevier ay. All rights reserved.

An efficient management scheme for updating redundant information in flash-based storage system Since flash memory has many attractive characteristics such as high performance, non-volatility, low power consumption and shock resistance, it has been widely used as a storage media in embedded and computer system environments. However, there are many shortcomings in flash memory such as potentially high I/O latency due to erase-before-write and poor durability due to limited erase cycles. To address these performance and reliability anomalies, many large-scale storage systems use redundancy-based parallel access schemes such as RAID techniques. However, such redundancy-based schemes incur high overhead due to generating and storing redundancy information, especially in flash-based storage systems. In this paper, we propose a novel and performance-effective approach using a redundancy-based data management scheme in flash storage, called Flash-aware Redundancy Array. The proposed technique not only reduces the redundancy management overhead by performing redundancy update operations during idle periods, but also provides a preventive mechanism to recover data from unexpected read errors occurring before such redundancy update operations finish. From the experiments, we found that the proposed technique improves flash-based storage systems by 19% in average execution time as compared to other redundancy-based approaches.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Conformant Planning via Model Checking . Conformant planning is the problem of nding a sequenceof actions that is guaranteed to achieve the goal for any possible initialstate and nondeterministic behavior of the planning domain. In this paperwe present a new approach to conformant planning. We propose analgorithm that returns the set of all conformant plans of minimal lengthif the problem admits a solution, otherwise it returns with failure. Ourwork is based on the planning via model checking paradigm, and relieson...

Generalizing the relaxed planning heuristic to non-linear tasks The relaxed planning heuristic is a prominent state-to-goal estimator function for domain-independent forward-chaining heuristic search and local search planning. It enriches the state-space traversal of almost all currently available suboptimal state-of-the-art planning systems. While current domain description languages allow general arithmetic expressions in precondition and effect lists, the heuristic has been devised for propositional, restricted, and linear tasks only. On the other hand, generalizations of the heuristic to non-linear tasks are of apparent need for modelling complex planning problems and a true necessity to validate software. Subsequently, this work proposes a solid extension to the estimate that can deal with non-linear preconditions and effects. It is derived based on an approximated plan construction with respect to intervals for variable assignments. For plan extraction, weakest preconditions are computed according to the assignment rule in Hoare's calculus.

Pushing Goal Derivation in DLP Computations dlv is a knowledge representation system, based on disjunctive logic programming, which offers front-ends to several advanced KR formalisms. This paper describes new techniques for the computation of answer sets of disjunctive logic programs, that have been developed and implemented in the dlv system. These techniques try to "push" the query goals in the process of model generation (query goals are often present either explicitly, like in planning and diagnosis, or implicitly in the form of integrity constraints). This way, a lot of useless models are discarded "a priori" and the computation converges rapidly toward the generation of the "right" answer set. A few preliminary benchmarks show dramatic efficiency gains due to the new techniques.

Planning with sensing, concurrency, and exogenous events: logical framework and implementation The focus of current research in cognitive robotics is both on the realization of sys- tems based on known formal settings and on the extension of previous formal approaches to account for features that play a signifl- cant role for autonomous robots, but have not yet received an adequate treatment. In this paper we adopt a formal framework de- rived from Propositional Dynamic Logics by exploiting their formal correspondence with Description Logics, and present an extension of such a framework obtained by introducing both concurrency on primitive actions and autoepistemic operators for explicitly repre- senting the robot's epistemic state. We show that the resulting formal setting allows for the representation of actions with context- dependent efiects, sensing actions, and con- current actions, and address both the pres- ence of exogenous events and the characteri- zation of the notion of executable plan in such a complex setting. Moreover, we present an implementation of this framework in a system which is capable of generating plans that are actually executed on mobile robots, and illus- trate the experimentation of such a system in the design and implementation of soccer players for the 1999 Robocup competition.

Open World Planning in the Situation Calculus We describe a forward reasoning planner for openworlds that uses domain specific information for pruningits search space, as suggested by (Bacchus &amp; Kabanza1996; 2000). The planner is written in the situationcalculus-based programming language GOLOG,and it uses a situation calculus axiomatization of theapplication domain. Given a sentence oe to prove, theplanner regresses it to an equivalent sentence oe 0 aboutthe initial situation, then invokes a theorem prover todetermine...

Conformant Planning via Heuristic Forward Search: A New Approach Conformant planning is the task of generating plans given un- certainty about the initial state and action effects, and with- out any sensing capabilities during plan execution. The plan should be successful regardless of which particular initial world we start from. It is well known that conformant plan- ning can be transformed into a search problem in belief space, the space whose elements are sets of possible worlds. We in- troduce a new representation of that search space, replacing the need to store sets of possible worlds with a need to rea- son about the effects of action sequences. The reasoning is done by deciding solvability of CNFs that capture the action sequence's semantics. Based on this approach, we extend the classical heuristic planning system FF to the conformant set- ting. The key to this extension is the introduction of approx- imative CNF reasoning in FF's heuristic function. Our ex- perimental evaluation shows Conformant-FF to be superior to the state-of-the-art conformant planners MBP, KACMBP, and GPT in a variety of benchmark domains.

Answer set programming and plan generation The idea of answer set programming is to represent a given computational problem by a logic program whose answer sets correspond to solutions, and then use an answer set solver, such as SMODELS or DLV, to find an answer set for this program. Applications of this method to planning are related to the line of research on the frame problem that started with the invention of formal nonmonotonic reasoning in 1980.

The FF planning system: fast plan generation through heuristic search We describe and evaluate the algorithmic techniques that are used in the FF planning system. Like the HSP system, FF relies on forward state space search, using a heuristic that estimates goal distances by ignoring delete lists. Unlike HSP's heuristic, our method does not assume facts to be independent. We introduce a novel search strategy that combines hill-climbing with systematic search, and we show how other powerful heuristic information can be extracted and used to prune the search space. FF was the most successful automatic planner at the recent AIPS-2000 planning competition. We review the results of the competition, give data for other benchmark domains, and investigate the reasons for the runtime performance of FF compared to HSP.

Reasoning about Complex Actions with Incomplete Knowledge: A Modal Approach In this paper we propose a modal approach for reasoning about dynamic domains in a logic programming setting. We present a logical framework for reasoning about actions in which modal inclusion axioms of the form 驴p0驴驴 驴 驴p1驴 驴p2驴 ... 驴pn驴 allow procedures to be defined for building complex actions from elementary actions. The language is able to handle knowledge producing actions as well as actions which remove information. Incomplete states are represented by means of epistemic operators and test actions can be used to check whether a fluent is true, false or undefined in a state. We give a non-monotonic solution for the frame problem by making use of persistency assumptions in the context of an abductive characterization. A goal directed proof procedure is defined, which allows reasoning about complex actions and generating conditional plans.

Planning with Reduced Operator Sets Classical propositional STRIPS planning is nothingbut the search for a path in the state-transition graphinduced by the operators in the planning problem.What makes the problem hard is the size and the sometimesadverse structure of this graph. We conjecturethat the search for a plan would be more efficient ifthere were only a small number of paths from the initialstate to the goal state. To verify this conjecture, wedefine the notion of reduced operator sets and describeways...

Planning with h+in theory and practice Many heuristic estimators for classical planning are based on the so-called delete relaxation, which ignores negative effects of planning operators. Ideally, such heuristics would compute the actual goal distance in the delete relaxation, i.e., the cost of an optimal relaxed plan, denoted by h+. However, current delete relaxation heuristics only provide (often inadmissible) estimates to h+ because computing the correct value is an NP-hard problem. In this work, we consider the approach of planning with the actual h+ heuristic from a theoretical and computational per- spective. In particular, we provide domain-dependent com- plexity results that classify some standard benchmark do- mains into ones where h+ can be computed efficiently and ones where computing h+ is NP-hard. Moreover, we study domain-dependent implementations of h+ which show that the h+ heuristic provides very informative heuristic estimates compared to other state-of-the-art heuristics.

The Astral Compendium For Protein Structure And Sequence Analysis The ASTRAL compendium provides several databases and tools to aid in the analysis of protein structures, particularly through the use of their sequences. The SPACI scores included in the system summarize the overall characteristics of a protein structure. A structural alignments database indicates residue equivalencies in superimposed protein domain structures, The PDB sequence-map files provide a linkage between the amino acid sequence of the molecule studied (SEQRES records in a database entry) and the sequence of the atoms experimentally observed in the structure (ATOM records). These maps are combined with information in the SCOP database to provide sequences of protein domains. Selected subsets of the domain database, with varying degrees of similarity measured in several different ways, are also available. ASTRAL may be accessed at http://astral.stanford.edu/.

Prefetching over a network: early experience with CTIP We discuss CTIP, an implementation of a network filesystem extension of the successful TIP informed prefetching and cache management system. Using a modified version of TIP in NFS client machines (and unmodified NFS servers). CTIP takes advantage of application-supplied hints that disclose the application's future read accesses. CTIP uses these hints to aggressively prefetch file data from an NFS file server and to make better local cache replacement decisions. This prefetching hides disk latency and exposes storage parallelism. Preliminary measurements that show CTIP can reduce execution time by a ratio comparable to that obtained with local TIP over a suite of I/O-intensive hinting applications. (For four disks, the reductions in execution time range from 17% to 69%). If local TIP execution requires that data first be loaded from remote storage into a local scratch area, then CTIP execution is significantly faster than the aggregate time for loading the data and executing. Additionally, our measurements show that the benefit of CTIP for hinting applications improves in the face of competition from other clients for server resources. We conclude with an analysis of the remaining problems with using unmodified NFS servers.

Improving Citation Polarity Classification With Product Reviews Recent work classifying citations in scientific literature has shown that it is possible to improve classification results with extensive feature engineering. While this result confirms that citation classification is feasible, there are two drawbacks to this approach: (i) it requires a large annotated corpus for supervised classification, which in the case of scientific literature is quite expensive; and (ii) feature engineering that is too specific to one area of scientific literature may not be portable to other domains, even within scientific literature. In this paper we address these two drawbacks. First, we frame citation classification as a domain adaptation task and leverage the abundant labeled data available in other domains. Then, to avoid over-engineering specific citation features for a particular scientific domain, we explore a deep learning neural network approach that has shown to generalize well across domains using unigram and bigram features. We achieve better citation classification results with this cross-domain approach than using in-domain classification.

A Resistive CAM Processing-in-Storage Architecture for DNA Sequence Alignment. A novel processing-in-storage (PRinS) architecture based on Resistive CAM (ReCAM) is described and proposed for Smith-Waterman (S-W) sequence alignment. The ReCAM PRinS massively parallel compare operation finds matching base pairs in a fixed number of cycles, regardless of sequence length. The ReCAM PRinS S-W algorithm is simulated and compared to FPGA, Xeon Phi, and GPU-based implementations, sh...

CUDAlign 4.0: Incremental Speculative Traceback for Exact Chromosome-Wide Alignment in GPU Clusters. This paper proposes and evaluates CUDAlign 4.0, a parallel strategy to obtain the optimal alignment of huge DNA sequences in multi-GPU platforms, using the exact Smith–Waterman (SW) algorithm. In the first phase of CUDAlign 4.0, a huge Dynamic Programming (DP) matrix is computed by multiple GPUs, which asynchronously communicate border elements to the right neighbor in order to find the optimal sc...

The FPGA-Based High-Performance Computer RIVYERA for Applications in Bioinformatics.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Logic Programming and Negation: A Survey. We survey here various approaches which were proposed to incorporate negation in logicprograms. We concentrate on the proof-theoretic and model-theoretic issues and the relationshipsbetween them.1991 Mathematics Subject Classification: 68Q40, 68T15.CR Categories: F.3.2., F.4.1, H.3.3, I.2.3.Keywords and Phrases: negation, general logic programs, non-monotonic reasoning.Notes. The work of the first author was partly supported by ESPRIT Basic Research Action6810 (Compulog 2). The work...

A sufficient condition for backtrack-bounded search Backtrack search is often used to solve constraint satisfaction problems. A relationship involving the structure of the constraints is described that provides a bound on the backtracking required to advance deeper into the backtrack tree. This analysis leads to upper bounds on the effort required for solution of a class of constraint satisfaction problems. The solutions involve a combination of relaxation preprocessing and backtrack search. The bounds are expressed in terms of the structure of the constraint connections. Specifically, the effort is shown to have a bound exponential in the size of the largest biconnected component of the constraint graph, as opposed to the size of the graph as a whole.

Convergence of a Nonconforming Multiscale Finite Element Method The multiscale finite element method (MsFEM) [T. Y. Hou, X. H. Wu, and Z. Cai, Math. Comp., 1998, to appear; T. Y. Hou and X. H. Wu, J. Comput. Phys., 134 (1997), pp. 169--189] has been introduced to capture the large scale solutions of elliptic equations with highly oscillatory coefficients. This is accomplished by constructing the multiscale base functions from the local solutions of the elliptic operator. Our previous study reveals that the leading order error in this approach is caused by the ``resonant sampling,'' which leads to large error when the mesh size is close to the small scale of the continuous problem. Similar difficulty also arises in numerical upscaling methods. An oversampling technique has been introduced to alleviate this difficulty [T. Y. Hou and X. H. Wu, J. Comput. Phys., 134 (1997), pp. 169--189]. A consequence of the oversampling method is that the resulting finite element method is no longer conforming. Here we give a detailed analysis of the nonconforming error. Our analysis also reveals a new cell resonance error which is caused by the mismatch between the mesh size and the wavelength of the small scale. We show that the cell resonance error is of lower order. Our numerical experiments demonstrate that the cell resonance error is generically small and is difficult to observe in practice.

Efficient sparse coding algorithms Sparse coding provides a class of algorithms for finding succinct representations of stimuli; given only unlabeled input data, it discovers basis functions that cap- ture higher-level features in the data. However, finding sparse codes remains a very difficult computational problem. In this paper, we present efficient sparse coding algorithms that are based on iteratively solving two convex optimization problems: an L1-regularized least squares problem and an L2-constrained least squares problem. We propose novel algorithms to solve both of these optimiza- tion problems. Our algorithms result in a significant speedup for sparse coding, allowing us to learn larger sparse codes than possible with previously described algorithms. We apply these algorithms to natural images and demonstrate that the inferred sparse codes exhibit end-stopping and non-classical receptive field sur- round suppression and, therefore, may provide a partial explanation for these two phenomena in V1 neurons.

Synchronized Disk Interleaving A group of disks may be interleaved to speed up data transfers in a manner analogous to the speedup achieved by main memory interleaving. Conventional disks may be used for interleaving by spreading data across disks and by treating multiple disks as if they were a single one. Furthermore, the rotation of the interleaved disks may be synchronized to simplify control and also to optimize performance. In addition, check- sums may be placed on separate check-sum disks in order to improve reliability. In this paper, we study synchronized disk interleaving as a high-performance mass storage system architecture. The advantages and limitations of the proposed disk interleaving scheme are analyzed using the M/G/1 queueing model and compared to the conventional disk access mechanism.

Downward Separation Fails Catastrophically for Limited Nondeterminism Classes The $\beta$ hierarchy consists of classes $\beta_k={\rm NP}[logkn]\subseteq {\rm NP}$. Unlike collapses in the polynomial hierarchy and the Boolean hierarchy, collapses in the $\beta$ hierarchy do not seem to translate up, nor does closure under complement seem to cause the hierarchy to collapse. For any consistent set of collapses and separations of levels of the hierarchy that respects ${\rm P} = \beta_1\subseteq \beta_2\subseteq \cdots \subseteq {\rm NP}$, we can construct an oracle relative to which those collapses and separations hold; at the same time we can make distinct levels of the hierarchy closed under computation or not, as we wish. To give two relatively tame examples: for any $k \geq 1$, we construct an oracle relative to which \[ {\rm P} = \beta_{k} \neq \beta_{k+1} \neq \beta_{k+2} \neq \cdots \] and another oracle relative to which \[ {\rm P} = \beta_{k} \neq \beta_{k+1} = {\rm PSPACE}. \] We also construct an oracle relative to which $\beta_{2k} = \beta_{2k+1} \neq \beta_{2k+2}$ for all k.

Normal forms for answer sets programming Normal forms for logic programs under stable/answer set semantics are introduced. We argue that these forms can simplify the study of program properties, mainly consistency. The first normal form, called the kernel of the program, is useful for studying existence and number of answer sets. A kernel program is composed of the atoms which are undefined in the Well-founded semantics, which are those that directly affect the existence of answer sets. The body of rules is composed of negative literals only. Thus, the kernel form tends to be significantly more compact than other formulations. Also, it is possible to check consistency of kernel programs in terms of colorings of the Extended Dependency Graph program representation which we previously developed. The second normal form is called 3-kernel. A 3-kernel program is composed of the atoms which are undefined in the Well-founded semantics. Rules in 3-kernel programs have at most two conditions, and each rule either belongs to a cycle, or defines a connection between cycles. 3-kernel programs may have positive conditions. The 3-kernel normal form is very useful for the static analysis of program consistency, i.e. the syntactic characterization of existence of answer sets. This result can be obtained thanks to a novel graph-like representation of programs, called Cycle Graph which presented in the companion article Costantini (2004b).

A cost-benefit scheme for high performance predictive prefetching

Scheduling parallel I/O operations The I/O bottleneck in parallel computer systems has recently begun receiving increasing interest. Most attention has focused on improving the performance of I/O devices using fairly low-level parallelism in techniques such as disk striping and interleaving. Widely applicable solutions, however, will require an integrated approach which addresses the problem at multiple system levels, including applications, systems software, and architecture. We propose that within the context of such an integrated approach, scheduling parallel I/O operations will become increasingly attractive and can potentially provide substantial performance benefits.We describe a simple I/O scheduling problem and present approximate algorithms for its solution. The costs of using these algorithms in terms of execution time, and the benefits in terms of reduced time to complete a batch of I/O operations, are compared with the situations in which no scheduling is used, and in which an optimal scheduling algorithm is used. The comparison is performed both theoretically and experimentally. We have found that, in exchange for a small execution time overhead, the approximate scheduling algorithms can provide substantial improvements in I/O completion times.

Learning Topic Representation For Smt With Neural Networks Statistical Machine Translation (SMT) usually utilizes contextual information to disambiguate translation candidates. However, it is often limited to contexts within sentence boundaries, hence broader topical information cannot be leveraged. In this paper, we propose a novel approach to learning topic representation for parallel data using a neural network architecture, where abundant topical contexts are embedded via topic relevant monolingual data. By associating each translation rule with the topic representation, topic relevant rules are selected according to the distributional similarity with the source text during SMT decoding. Experimental results show that our method significantly improves translation accuracy in the NIST Chinese-to-English translation task compared to a state-of-the-art baseline.

MIND: A black-box energy consumption model for disk arrays Energy consumption is becoming a growing concern in data centers. Many energy-conservation techniques have been proposed to address this problem. However, an integrated method is still needed to evaluate energy efficiency of storage systems and various power conservation techniques. Extensive measurements of different workloads on storage systems are often very time-consuming and require expensive equipments. We have analyzed changing characteristics such as power and performance of stand-alone disks and RAID arrays, and then defined MIND as a black box power model for RAID arrays. MIND is devised to quantitatively measure the power consumption of redundant disk arrays running different workloads in a variety of execution modes. In MIND, we define five modes (idle, standby, and several types of access) and four actions, to precisely characterize power states and changes of RAID arrays. In addition, we develop corresponding metrics for each mode and action, and then integrate the model and a measurement algorithm into a popular trace tool - blktrace. With these features, we are able to run different IO traces on large-scale storage systems with power conservation techniques. Accurate energy consumption and performance statistics are then collected to evaluate energy efficiency of storage system designs and power conservation techniques. Our experiments running both synthetic and real-world workloads on enterprise RAID arrays show that MIND can estimate power consumptions of disk arrays with an error rate less than 2%.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

Parameterized complexity for the database theorist

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Planning-based configuration and management of distributed systems The configuration and runtime management of distributed systems is often complex due to the presence of a large number of configuration options and dependencies between interacting sub-systems. Inexperienced users usually choose default configurations because they are not aware of the possible configurations and/or their effect on the systems' operation. In doing so, they are unable to take advantage of the potentially wide range of system capabilities. Furthermore, managing inter-dependent sub-systems frequently involves performing a set of actions to get the overall system to the desired final state. In this paper, we propose a new approach for configuring and managing distributed systems based on AI planning. We use a goal-driven, tag-based user interaction paradigm to shield users from the complexities of configuring and managing systems. The key idea behind our approach is to package different configuration options and system management actions into reusable modules that can be automatically composed into workflows based on the user's goals. It also allows capturing the inter-dependencies between different configuration options, management actions and system states. We evaluate our approach in a case study involving three interdependent sub-systems. Our initial experiences indicate that this planning-based approach holds great promise in simplifying configuration and management tasks.

A planning based approach to failure recovery in distributed systems Failure recovery in distributed systems poses a difficult challenge because of the requirement for high availability. Failure scenarios are usually unpredictable so they can not easily be foreseen. In this research we propose a planning based approach to failure recovery. This approach automates failure recovery by capturing the state after failure, defining an acceptable recovered state as a goal and applying planning to get from the initial state to the goal state. By using planning, this approach can recover from a variety of failed states and reach any of several acceptable states: from minimal functionality to complete recovery.

SAP speaks PDDL: exploiting a software-engineering model for planning in business process management Planning is concerned with the automated solution of action sequencing problems described in declarative languages giving the action preconditions and effects. One important application area for such technology is the creation of new processes in Business Process Management (BPM), which is essential in an ever more dynamic business environment. A major obstacle for the application of Planning in this area lies in the modeling. Obtaining a suitable model to plan with - ideally a description in PDDL, the most commonly used planning language - is often prohibitively complicated and/or costly. Our core observation in this work is that this problem can be ameliorated by leveraging synergies with model-based software development. Our application at SAP, one of the leading vendors of enterprise software, demonstrates that even one-to-one model re-use is possible. The model in question is called Status and Action Management (SAM). It describes the behavior of Business Objects (BO), i.e., large-scale data structures, at a level of abstraction corresponding to the language of business experts. SAM covers more than 400 kinds of BOs, each of which is described in terms of a set of status variables and how their values are required for, and affected by, processing steps (actions) that are atomic from a business perspective. SAM was developed by SAP as part of a major model-based software engineering effort. We show herein that one can use this same model for planning, thus obtaining a BPM planning application that incurs no modeling overhead at all. We compile SAM into a variant of PDDL, and adapt an off-the-shelf planner to solve this kind of problem. Thanks to the resulting technology, business experts may create new processes simply by specifying the desired behavior in terms of status variable value changes: effectively, by describing the process in their own language.

Contingent planning with goal preferences The importance of the problems of contingent planning with actions that have non-deterministic effects and of planning with goal preferences has been widely recognized, and several works address these two problems separately. However, combining conditional planning with goal preferences adds some new difficulties to the problem. Indeed, even the notion of optimal plan is far from trivial, since plans in nondeterministic domains can result in several different behaviors satisfying conditions with different preferences. Planning for optimal conditional plans must therefore take into account the different behaviors, and conditionally search for the highest preference that can be achieved. In this paper, we address this problem. We formalize the notion of optimal conditional plan, and we describe a correct and complete planning algorithm that is guaranteed to find optimal solutions. We implement the algorithm using BDD-based techniques, and show the practical potentialities of our approach through a preliminary experimental evaluation.

Compiling uncertainty away in conformant planning problems with bounded width Conformant planning is the problem of finding a sequence of actions for achieving a goal in the presence of uncertainty in the initial state or action effects. The problem has been approached as a path-finding problem in belief space where good belief representations and heuristics are critical for scaling up. In this work, a different formulation is introduced for conformant problems with deterministic actions where they are automatically converted into classical ones and solved by an off-the-shelf classical planner. The translation maps literals L and sets of assumptions t about the initial situation, into new literals KL/t that represent that L must be true if t is initially true. We lay out a general translation scheme that is sound and establish the conditions under which the translation is also complete. We show that the complexity of the complete translation is exponential in a parameter of the problem called the conformant width, which for most benchmarks is bounded. The planner based on this translation exhibits good performance in comparison with existing planners, and is the basis for T0, the best performing planner in the Conformant Track of the 2006 International Planning Competition.

Complexity Results For Sas(+) Planning We have previously reported a number of tractable planning problems defined in the SAS(+) formalism. This article complements these results by providing a complete map over the complexity of SAS(+) planning under all combinations of the previously considered restrictions. We analyze the complexity of both finding a minimal plan and finding any plan. In contrast to other complexity surveys of planning, we study not only the complexity of the decision problems but also the complexity of the generation problems. We prove that the SAS(+)-PUS problem is the maximal tractable problem under the restrictions we have considered if we want to generate minimal plans. If we are satisfied with any plan, then we can generalize further to the SAS(+)-US problem, which we prove to be the maximal tractable problem in this case.

Local Search Topology in Planning Benchmarks: A Theoretical Analysis. Many state-of-the-art heuristic planners derive their heuristic function by relaxing the planning task at hand, where the relaxation is to assume that all delete lists are empty. The success of such planners on many of the current benchmarks suggests that in those task’s state spaces relaxed goal distances yield a heuristic function of high quality. Recent work has revealed empirical evidence confirming this intuition, stating several hypotheses about the local search topology of the current benchmarks, concerning the non-existence of dead ends and of local minima, as well as a limited maximal distance to exits on benches. Investigating a large range of planning domains, we prove that the above hypotheses do in fact hold true for the majority of the current benchmarks. This explains the recent success of heuristic planners. Specifically, it follows that FF’s search algorithm, using an idealized heuristic function, is polynomial in (at least) eight commonly used benchmark domains. Our proof methods shed light on what the structural reasons are behind the topological phenomena, giving hints on how these phenomena might be automatically recognizable.

Logic Programming and Negation: A Survey. We survey here various approaches which were proposed to incorporate negation in logicprograms. We concentrate on the proof-theoretic and model-theoretic issues and the relationshipsbetween them.1991 Mathematics Subject Classification: 68Q40, 68T15.CR Categories: F.3.2., F.4.1, H.3.3, I.2.3.Keywords and Phrases: negation, general logic programs, non-monotonic reasoning.Notes. The work of the first author was partly supported by ESPRIT Basic Research Action6810 (Compulog 2). The work...

Learning Face Representation from Scratch. Pushing by big data and deep convolutional neural network (CNN), the performance of face recognition is becoming comparable to human. Using private large scale training datasets, several groups achieve very high performance on LFW, i.e., 97% to 99%. While there are many open source implementations of CNN, none of large scale face dataset is publicly available. The current situation in the field of face recognition is that data is more important than algorithm. To solve this problem, this paper proposes a semi-automatical way to collect face images from Internet and builds a large scale dataset containing about 10,000 subjects and 500,000 images, called CASIAWebFace. Based on the database, we use a 11-layer CNN to learn discriminative representation and obtain state-of-theart accuracy on LFW and YTF. The publication of CASIAWebFace will attract more research groups entering this field and accelerate the development of face recognition in the wild.

The LOCKSS peer-to-peer digital preservation system The LOCKSS project has developed and deployed in a world-wide test a peer-to-peer system for preserving access to journals and other archival information published on the Web. It consists of a large number of independent, low-cost, persistent Web caches that cooperate to detect and repair damage to their content by voting in “opinion polls.” Based on this experience, we present a design for and simulations of a novel protocol for voting in systems of this kind. It incorporates rate limitation and intrusion detection to ensure that even some very powerful adversaries attacking over many years have only a small probability of causing irrecoverable damage before being detected.

LRFU: A Spectrum of Policies that Subsumes the Least Recently Used and Least Frequently Used Policies Efficient and effective buffering of disk blocks in main memory is critical for better file system performance due to a wide speed gap between main memory and hard disks. In such a buffering system, one of the most important design decisions is the block replacement policy that determines which disk block to replace when the buffer is full. In this paper, we show that there exists a spectrum of block replacement policies that subsumes the two seemingly unrelated and independent Least Recently Used (LRU) and Least Frequently Used (LFU) policies. The spectrum is called the LRFU (Least Recently/Frequently Used) policy and is formed by how much more weight we give to the recent history than to the older history. We also show that there is a spectrum of implementations of the LRFU that again subsumes the LRU and LFU implementations. This spectrum is again dictated by how much weight is given to recent and older histories and the time complexity of the implementations lies between O(1) (the time complexity of LRU) and {\rm O}(\log_2 n) (the time complexity of LFU), where n is the number of blocks in the buffer. Experimental results from trace-driven simulations show that the performance of the LRFU is at least competitive with that of previously known policies for the workloads we considered.

Low-density parity-check matrices for coding of correlated sources Linear codes for a coding problem of correlated sources are considered. It is proved that we can construct codes by using low-density parity-check (LDPC) matrices with maximum-likelihood (or typical set) decoding. As applications of the above coding problem, a construction of codes is presented for multiple-access channel with correlated additive noises and a coding theorem of parity-check codes for general channels is proved.

Representing the Process Semantics in the Event Calculus In this paper we shall present a translation of the process semantics [5] to the event calculus. The aim is to realize a method of integrating high-level semantics with logical calculi to reason about continuous change. The general translation rules and the soundness and completeness theorem of the event calculus with respect to the process semantics are main technical results of this paper.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Protection in the Hydra Operating System This paper describes the capability based protection mechanisms provided by the Hydra Operating System Kernel. These mechanisms support the construction of user-defined protected subsystems, including file and directory subsystems, which do not therefore need to be supplied directly by Hydra. In addition, we discuss a number of well known protection problems, including Mutual Suspicion, Confinement and Revocation, and we present the mechanisms that Hydra supplies in order to solve them.

Fine-Grained Mobility in the Emerald System (Extended Abstract)

The Narrowing Gap Between Language Systems and Operating Systems

A probe-based monitoring scheme for an object-oriented distributed operating system

Fault Tolerant Computing in Object Based Distributed Operating Systems

Wave Scheduling: Distributed Allocation of Task Forces in Network Computers

A principle for resilient sharing of distributed resources A technique is described which permits distributed resources to be shared (services to be offered) in a resilient manner. The essence of the technique is to a priori declare one of the server hosts primary and the others backups. Any of the servers can perform the primary duties. Thus the role of primary can migrate around the set of servers. The concept of n-host resiliency is introduced and the error detection and recovery schemes for two-host resiliency are presented. The single primary, multiple backup technique for resource sharing is shown to have minimal delay. In the general case, this is superior to multiple primary techniques.

A Case for Fault-Tolerant Memory for Transaction Processing

Read Optimized File System Designs: A Performance Evaluation This paper presents a performance comparison of several file system allocation policies. The file systems are designed to provide high bandwidth between disks and main memory by taking advantage of parallelism in an underlying disk array, catering to large units of transfer, and minimizing the bandwidth dedicated to the transfer of meta data. All of the file systems described use a mul- tiblock allocation strategy which allows both large and small files to be allocated efficiently. Simulation results show that these multiblock policies result in systems that are able to utilize a large percentage of the underlying disk bandwidth; more than 90% in sequential cases. As general purpose systems are called upon to support more data intensive applications such as databases and super- computing, these policies offer an opportunity to provide superior performance to a larger class of users.

Flexible buffer allocation based on marginal gazns Previous works on buflcx allocation are based f~il$lwr exclusively on the availability of buffers at r{ll)timc or on the access pat t eras of queries. In this paper We p repose a unified approach for buffer allocation in which both of these considerations are taken into accou at. Our approach is based on the notion of marginal y~~ins which specify the expected reduction cm page faults in allocating extra buffers to a query. Simulation results show that our approach is promising, and allocation algorithms based on marginal gains perform cousidwably better than existing on’es.

DRPM: dynamic speed control for power management in server class disks A large portion of the power budget in server environments goes into the I/O subsystem - the disk array in particular. Traditional approaches to disk power management involve completely stopping the disk rotation, which can take a considerable amount of time, making them less useful in cases where idle times between disk requests may not be long enough to outweigh the overheads. This paper presents a new approach called DRPM to modulate disk speed (RPM) dynamically, and gives a practical implementation to exploit this mechanism. Extensive simulations with different workload and hardware parameters show that DRPM can provide significant energy savings without compromising much on performance. This paper also discusses practical issues when implementing DRPM on server disks.

Deep Gaussian Processes In this paper we introduce deep Gaussian process (GP) models. Deep GPs are a deep belief network based on Gaussian process mappings. The data is modeled as the output of a multivariate GP. The inputs to that Gaussian process are then governed by another GP. A single layer model is equivalent to a standard GP or the GP latent variable model (GP-LVM). We perform inference in the model by approximate variational marginalization. This results in a strict lower bound on the marginal likelihood of the model which we use for model selection (number of layers and nodes per layer). Deep belief networks are typically applied to relatively large data sets using stochastic gradient descent for optimization. Our fully Bayesian treatment allows for the application of deep models even when data is scarce. Model selection by our variational bound shows that a five layer hierarchy is justified even when modelling a digit data set containing only 150 examples.

The Complexity of Read-Once Resolution We investigate the complexity of deciding whether a propositional formula has a read-once resolution proof. We give a new and general proof of Iwama–Miynano's theorem which states that the problem whether a formula has a read-once resolution proof is iNP-complete. Moreover, we show for fixed ik&ges;2 that the additional restriction that in each resolution step one of the parent clauses is a ik-clause preserves the iNP-completeness. If we demand that the formulas are minimal unsatisfiable and read-once refutable then the problem remains iNP-complete. For the subclasses iMU(ik) of minimal unsatisfiable formulas we present a pol-time algorithm deciding whether a iMU(ik)-formula has a read-once resolution proof. Furthermore, we show that the problems whether a formula contains a iMU(ik)-subformula or a read-once refutable iMU(ik)-subformula are iNP-complete.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

A Theory of Cheap Control in Embodied Systems. We present a framework for designing cheap control architectures of embodied agents. Our derivation is guided by the classical problem of universal approximation, whereby we explore the possibility of exploiting the agent's embodiment for a new and more efficient universal approximation of behaviors generated by sensorimotor control. This embodied universal approximation is compared with the classical non-embodied universal approximation. To exemplify our approach, we present a detailed quantitative case study for policy models defined in terms of conditional restricted Boltzmann machines. In contrast to non-embodied universal approximation, which requires an exponential number of parameters, in the embodied setting we are able to generate all possible behaviors with a drastically smaller model, thus obtaining cheap universal approximation. We test and corroborate the theory experimentally with a six-legged walking machine. The experiments indicate that the controller complexity predicted by our theory is close to the minimal sufficient value, which means that the theory has direct practical implications.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Learning Topic Representation For Smt With Neural Networks Statistical Machine Translation (SMT) usually utilizes contextual information to disambiguate translation candidates. However, it is often limited to contexts within sentence boundaries, hence broader topical information cannot be leveraged. In this paper, we propose a novel approach to learning topic representation for parallel data using a neural network architecture, where abundant topical contexts are embedded via topic relevant monolingual data. By associating each translation rule with the topic representation, topic relevant rules are selected according to the distributional similarity with the source text during SMT decoding. Experimental results show that our method significantly improves translation accuracy in the NIST Chinese-to-English translation task compared to a state-of-the-art baseline.

Towards Self-configuration in Autonomic Electronic Institutions Electronic institutions (EIs) have been proposed as a means of regulating open agent societies. EIs define the rules of the game in agent societies by fixing what agents are permitted and forbidden to do and under what circumstances. And yet, there is the need for EIs to adapt their regulations to comply with their goals despite coping with varying populations of self-interested agents. In this paper we focus on the extension of EIs with autonomic capabilities to allow them to yield a dynamical answer to changing circumstances through the adaptation of their norms.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

Parameterized complexity for the database theorist

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

RODEO: Robust DE-aliasing autoencOder for real-time medical image reconstruction. In this work we address the problem of real-time dynamic medical (MRI and X-Ray CT) image reconstruction from parsimonious samples (Fourier frequency space for MRI and sinogram/tomographic projections for CT). Today the de facto standard for such reconstruction is compressed sensing (CS). CS produces high quality images (with minimal perceptual loss); but such reconstructions are time consuming, requiring solving a complex optimization problem. In this work we propose to ‘learn’ the reconstruction from training samples using an autoencoder. Our work is based on the universal function approximation capacity of neural networks. The training time for the autoencoder is large, but is offline and hence does not affect performance during operation. During testing/operation, our method requires only a few matrix vector products and hence is significantly faster than CS based methods. In fact, for MRI it is fast enough for real-time reconstruction (the images are reconstructed as fast as they are acquired) with only slight degradation of image quality; for CT our reconstruction speed is slightly slower than required for real-time reconstruction. However, in order to make the autoencoder suitable for our problem, we depart from the standard Euclidean norm cost function of autoencoders and use a robust l1-norm instead. The ensuing problem is solved using the Split Bregman method.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Learning Topic Representation For Smt With Neural Networks Statistical Machine Translation (SMT) usually utilizes contextual information to disambiguate translation candidates. However, it is often limited to contexts within sentence boundaries, hence broader topical information cannot be leveraged. In this paper, we propose a novel approach to learning topic representation for parallel data using a neural network architecture, where abundant topical contexts are embedded via topic relevant monolingual data. By associating each translation rule with the topic representation, topic relevant rules are selected according to the distributional similarity with the source text during SMT decoding. Experimental results show that our method significantly improves translation accuracy in the NIST Chinese-to-English translation task compared to a state-of-the-art baseline.

Computational diagnosis of skin lesions from dermoscopic images using combined features There has been an alarming increase in the number of skin cancer cases worldwide in recent years, which has raised interest in computational systems for automatic diagnosis to assist early diagnosis and prevention. Feature extraction to describe skin lesions is a challenging research area due to the difficulty in selecting meaningful features. The main objective of this work is to find the best combination of features, based on shape properties, colour variation and texture analysis, to be extracted using various feature extraction methods. Several colour spaces are used for the extraction of both colour- and texture-related features. Different categories of classifiers were adopted to evaluate the proposed feature extraction step, and several feature selection algorithms were compared for the classification of skin lesions. The developed skin lesion computational diagnosis system was applied to a set of 1104 dermoscopic images using a cross-validation procedure. The best results were obtained by an optimum-path forest classifier with very promising results. The proposed system achieved an accuracy of 92.3%, sensitivity of 87.5% and specificity of 97.1% when the full set of features was used. Furthermore, it achieved an accuracy of 91.6%, sensitivity of 87% and specificity of 96.2%, when 50 features were selected using a correlation-based feature selection algorithm.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Learning Topic Representation For Smt With Neural Networks Statistical Machine Translation (SMT) usually utilizes contextual information to disambiguate translation candidates. However, it is often limited to contexts within sentence boundaries, hence broader topical information cannot be leveraged. In this paper, we propose a novel approach to learning topic representation for parallel data using a neural network architecture, where abundant topical contexts are embedded via topic relevant monolingual data. By associating each translation rule with the topic representation, topic relevant rules are selected according to the distributional similarity with the source text during SMT decoding. Experimental results show that our method significantly improves translation accuracy in the NIST Chinese-to-English translation task compared to a state-of-the-art baseline.

Predicting durability in DHTs using Markov chains We consider the problem of data durability in low- bandwidth large-scale distributed storage systems. Given the limited bandwidth between replicas, these systems suf- fer from long repair times after a hard disk crash, mak- ing them vulnerable to data loss when several replicas fail within a short period of time. Recent work has suggested that the probability of data loss can be predicted by model- ing the number of live replicas using a Markov chain. This, in turn, can then be used to determine the number of replicas necessary to keep the loss probability under a given desired value. Previous authors have suggested that the model parame- ters can be estimated using an expression that is constant or linear on the number of replicas. Our simulations, however, show that neither is correct, as these parameter values grow sublinearly with the number of replicas. Moreover, we show that using a linear expression will result in the probabilit y of data loss being underestimated, while the constant expres- sion will produce a significant overestimation. Finally, we provide an empirical expression that yields a good approxi- mation of the sublinear parameter values. Our work can be viewed as a first step towards finding more accurate models to predict the durability of this type of systems.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

Parameterized complexity for the database theorist

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Novel Multisample Scheme for Inferring Phylogenetic Markers from Whole Genome Tumor Profiles Computational cancer phylogenetics seeks to enumerate the temporal sequences of aberrations in tumor evolution, thereby delineating the evolution of possible tumor progression pathways, molecular subtypes, and mechanisms of action. We previously developed a pipeline for constructing phylogenies describing evolution between major recurring cell types computationally inferred from whole-genome tumor profiles. The accuracy and detail of the phylogenies, however, depend on the identification of accurate, high-resolution molecular markers of progression, i.e., reproducible regions of aberration that robustly differentiate different subtypes and stages of progression. Here, we present a novel hidden Markov model (HMM) scheme for the problem of inferring such phylogenetically significant markers through joint segmentation and calling of multisample tumor data. Our method classifies sets of genome-wide DNA copy number measurements into a partitioning of samples into normal (diploid) or amplified at each probe. It differs from other similar HMM methods in its design specifically for the needs of tumor phylogenetics, by seeking to identify robust markers of progression conserved across a set of copy number profiles. We show an analysis of our method in comparison to other methods on both synthetic and real tumor data, which confirms its effectiveness for tumor phylogeny inference and suggests avenues for future advances.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Parallel object allocation via user-specified directives: a case study in traffic simulation Predefined and automatic approaches to allocation cannot always achieve satisfactory results, due to the very different behaviours in the dynamic need of resources of parallel applications. This paper presents the approach adopted in the Parallel Objects (PO) environment to let users drive object allocation in parallel/distributed architectures. A set of high-level directives permits users to specify the allocation needs of application objects; a dynamic load-balancing tool – part of the environment run-time support – uses the user-level allocation directives to tune its behaviour. This paper presents the PO implementation of an application example in the field of traffic simulation. The goal is to show the ease of use and the flexibility of the allocation directives, together with the effectiveness of the approach in improving the performances of dynamic object-oriented parallel applications.

HPO: a programming environment for object-oriented metacomputing Metacomputing is an emergent paradigm that makes possible to d istribute applications over a h eterogeneous set of computing systems to exploit all available resources. The paper presents the HPO environment f or object-oriented metacomputing. The HPO programming model i s based on the object-oriented pa radigm and defines architecture-independent and portable applications. The HPO support makes it possible to distribute applications over a n etwork of heterogeneous architectures. The paper describes this approach via several examples and evaluates its achieved performances.

Distributed, object-based programming systems The development of distributed operating systems and object-based programming languages makes possible an environment in which programs consisting of a set of interacting modules, or objects, may execute concurrently on a collection of loosely coupled processors. An object-based programming language encourages a methodology for designing and creating a program as a set of autonomous components, whereas a distributed operating system permits a collection of workstations or personal computers to be treated as a single entity. The amalgamation of these two concepts has resulted in systems that shall be referred to as distributed, object-based programming systems. This paper discusses issues in the design and implementation of such systems. Following the presentation of fundamental concepts and various object models, issues in object management, object interaction management, and physical resource management are discussed. Extensive examples are drawn from existing systems.

Extended stable semantics for normal and disjunctive programs

The nature of statistical learning theory~. First Page of the Article

A machine program for theorem-proving The programming of a proof procedure is discussed in connection with trial runs and possible improvements.

An Introduction to Least Commitment Planning Recent developments have clarified the process of generating partially ordered, partially specified sequences of actions whose execution will achieve an agent's goal. This article summarizes a progression of least commitment planners, starting with one that handles the simple STRIPS representation and ending with UCOPOP a planner that manages actions with disjunctive precondition, conditional effects, and universal quantification over dynamic universes. Along the way, I explain how Chapman's formulation of the modal truth criterion is misleading and why his NP-completeness result for reasoning about plans with conditional effects does not apply to UCOPOP.

Equilibria and steering laws for planar formations This paper presents a Lie group setting for the problem of control of formations, as a natural outcome of the analysis of a planar two-vehicle formation control law. The vehicle trajectories are described using the planar Frenet–Serret equations of motion, which capture the evolution of both the vehicle position and orientation for unit-speed motion subject to curvature (steering) control. The set of all possible (relative) equilibria for arbitrary G-invariant curvature controls is described (where G=SE(2) is a symmetry group for the control law), and a global convergence result for the two-vehicle control law is proved. An n-vehicle generalization of the two-vehicle control law is also presented, and the corresponding (relative) equilibria for the n-vehicle problem are characterized. Work is on-going to discover stability and convergence results for the n-vehicle problem.

Pushing the envelope: planning, propositional logic, and stochastic search Planning is a notoriously hard combinatorial search problem. In many interesting domains, current planning algorithms fail to scale up gracefully. By combining a general, stochastic search algorithm and appropriate problem encodings based on propositional logic, we are able to solve hard planning problems many times faster than the best current planning systems. Although stochastic methods have been shown to be very effective on a wide range of scheduling problems, this is the first demonstration of its power on truly challenging classical planning instances. This work also provides a new perspective on representational issues in planning.

A Completeness Result for SLDNF-Resolution Because of the possibility of floundering and infinite derivations, SLDNF-resolution is, in general, not complete. The classical approach [17] to get a completeness result is to restrict the attention to normal programs P and normal goals G, such that P or {G} is allowed and P is hierarchical. Unfortunately, the class of all normal programs and all normal goals meeting these requirements is not powerful enough to be of great practical importance. But after refining the concept of allowedness by taking modes [12] into account, we can broaden the notion of a hierarchical program, and thereby define a subclass of the class of normal programs and normal goals which is powerful enough to compute all primitive recursive functions without losing the completeness of SLDNF-resolution.

A logic programming approach to knowledge-state planning: Semantics and complexity We propose a new declarative planning language, called K, which is based on principles and methods of logic programming. In this language, transitions between states of knowledge can be described, rather than transitions between completely described states of the world, which makes the language well suited for planning under incomplete knowledge. Furthermore, our formalism enables the use of default principles in the planning process by supporting negation as failure. Nonetheless, K also supports the representation of transitions between states of the world (i.e., states of complete knowledge) as a special case, which shows that the language is very flexible. As we demonstrate on particular examples, the use of knowledge states may allow for a natural and compact problem representation. We then provide a thorough analysis of the computational complexity of K, and consider different planning problems, including standard planning and secure planning (also known as conformant planning) problems. We show that these problems have different complexities under various restrictions, ranging from NP to NEXPTIME in the propositional case. Our results form the theoretical basis for the DLVk system, which implements the language K on top of the DLV logic programming system.

iSAM: Incremental Smoothing and Mapping In this paper, we present incremental smoothing and mapping (iSAM), which is a novel approach to the simultaneous localization and mapping problem that is based on fast incremental matrix factorization. iSAM provides an efficient and exact solution by updating a QR factorization of the naturally sparse smoothing information matrix, thereby recalculating only those matrix entries that actually change. iSAM is efficient even for robot trajectories with many loops as it avoids unnecessary fill-in in the factor matrix by periodic variable reordering. Also, to enable data association in real time, we provide efficient algorithms to access the estimation uncertainties of interest based on the factored information matrix. We systematically evaluate the different components of iSAM as well as the overall algorithm using various simulated and real-world datasets for both landmark and pose-only settings.

When Multivariate Forecasting Meets Unsupervised Feature Learning - Towards a Novel Anomaly Detection Framework for Decision Support. Many organizations adopt information technologies to make intelligent decisions during operations. Time-series data plays a crucial role in supporting such decision making processes. Though current studies on time-series based decision making provide reasonably well results, the anomaly detection essence underling most of the scenarios and the plenitude of unlabeled data are largely overlooked and left unexplored. We argue that by using multivariate forecasting and unsupervised feature learning, these two important research gaps could be filled. We carried out two experiments in this study to testify our approach and the results showed that decision support performance was significantly improved. We also proposed a novel framework to integrate the two methods so that our approach may be generalized to a larger problem domain. We discussed the advantages, the limitations and the future work of our study. Both practical and theoretical contributions were also discussed in the paper. © 2012 by the AIS/ICIS Administrative Office All rights reserved.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Neural Networks and Particle Swarm Optimization for Function Approximation in Tri-SWACH Hull Design Tri-SWACH is a novel multihull ship design that is well suited to a wide range of industrial, commercial, and military applications, but which because of its novelty has few experimental studies on which to base further development work. Using a new form of particle swarm optimization that incorporates a strong element of stochastic search, Breeding PSO, it is shown it is possible to use multilayer nets to predict resistance functions for Tri-SWACH hullforms, including one function, the Residual Resistance Coefficient, which was found intractable with previously explored neural network training methods.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Learning Topic Representation For Smt With Neural Networks Statistical Machine Translation (SMT) usually utilizes contextual information to disambiguate translation candidates. However, it is often limited to contexts within sentence boundaries, hence broader topical information cannot be leveraged. In this paper, we propose a novel approach to learning topic representation for parallel data using a neural network architecture, where abundant topical contexts are embedded via topic relevant monolingual data. By associating each translation rule with the topic representation, topic relevant rules are selected according to the distributional similarity with the source text during SMT decoding. Experimental results show that our method significantly improves translation accuracy in the NIST Chinese-to-English translation task compared to a state-of-the-art baseline.

Deep neural network with weight sparsity control and pre-training extracts hierarchical features and enhances classification performance: Evidence from whole-brain resting-state functional connectivity patterns of schizophrenia. Functional connectivity (FC) patterns obtained from resting-state functional magnetic resonance imaging data are commonly employed to study neuropsychiatric conditions by using pattern classifiers such as the support vector machine (SVM). Meanwhile, a deep neural network (DNN) with multiple hidden layers has shown its ability to systematically extract lower-to-higher level information of image and speech data from lower-to-higher hidden layers, markedly enhancing classification accuracy. The objective of this study was to adopt the DNN for whole-brain resting-state FC pattern classification of schizophrenia (SZ) patients vs. healthy controls (HCs) and identification of aberrant FC patterns associated with SZ. We hypothesized that the lower-to-higher level features learned via the DNN would significantly enhance the classification accuracy, and proposed an adaptive learning algorithm to explicitly control the weight sparsity in each hidden layer via L1-norm regularization. Furthermore, the weights were initialized via stacked autoencoder based pre-training to further improve the classification performance. Classification accuracy was systematically evaluated as a function of (1) the number of hidden layers/nodes, (2) the use of L1-norm regularization, (3) the use of the pre-training, (4) the use of framewise displacement (FD) removal, and (5) the use of anatomical/functional parcellation. Using FC patterns from anatomically parcellated regions without FD removal, an error rate of 14.2% was achieved by employing three hidden layers and 50 hidden nodes with both L1-norm regularization and pre-training, which was substantially lower than the error rate from the SVM (22.3%). Moreover, the trained DNN weights (i.e., the learned features) were found to represent the hierarchical organization of aberrant FC patterns in SZ compared with HC. Specifically, pairs of nodes extracted from the lower hidden layer represented sparse FC patterns implicated in SZ, which was quantified by using kurtosis/modularity measures and features from the higher hidden layer showed holistic/global FC patterns differentiating SZ from HC. Our proposed schemes and reported findings attained by using the DNN classifier and whole-brain FC data suggest that such approaches show improved ability to learn hidden patterns in brain imaging data, which may be useful for developing diagnostic tools for SZ and other neuropsychiatric disorders and identifying associated aberrant FC patterns.

Handwritten Hangul recognition using deep convolutional neural networks In spite of the advances in recognition technology, handwritten Hangul recognition (HHR) remains largely unsolved due to the presence of many confusing characters and excessive cursiveness in Hangul handwritings. Even the best existing recognizers do not lead to satisfactory performance for practical applications and have much lower performance than those developed for Chinese or alphanumeric characters. To improve the performance of HHR, here we developed a new type of recognizers based on deep neural networks (DNNs). DNN has recently shown excellent performance in many pattern recognition and machine learning problems, but have not been attempted for HHR. We built our Hangul recognizers based on deep convolutional neural networks and proposed several novel techniques to improve the performance and training speed of the networks. We systematically evaluated the performance of our recognizers on two public Hangul image databases, SERI95a and PE92. Using our framework, we achieved a recognition rate of 95.96 % on SERI95a and 92.92 % on PE92. Compared with the previous best records of 93.71 % on SERI95a and 87.70 % on PE92, our results yielded improvements of 2.25 and 5.22 %, respectively. These improvements lead to error reduction rates of 35.71 % on SERI95a and 42.44 % on PE92, relative to the previous lowest error rates. Such improvement fills a significant portion of the large gap between practical requirement and the actual performance of Hangul recognizers.

3D Mesh Labeling via Deep Convolutional Neural Networks This article presents a novel approach for 3D mesh labeling by using deep Convolutional Neural Networks (CNNs). Many previous methods on 3D mesh labeling achieve impressive performances by using predefined geometric features. However, the generalization abilities of such low-level features, which are heuristically designed to process specific meshes, are often insufficient to handle all types of meshes. To address this problem, we propose to learn a robust mesh representation that can adapt to various 3D meshes by using CNNs. In our approach, CNNs are first trained in a supervised manner by using a large pool of classical geometric features. In the training process, these low-level features are nonlinearly combined and hierarchically compressed to generate a compact and effective representation for each triangle on the mesh. Based on the trained CNNs and the mesh representations, a label vector is initialized for each triangle to indicate its probabilities of belonging to various object parts. Eventually, a graph-based mesh-labeling algorithm is adopted to optimize the labels of triangles by considering the label consistencies. Experimental results on several public benchmarks show that the proposed approach is robust for various 3D meshes, and outperforms state-of-the-art approaches as well as classic learning algorithms in recognizing mesh labels.

Change Detection in Synthetic Aperture Radar Images Based on Deep Neural Networks This paper presents a novel change detection approach for synthetic aperture radar images based on deep learning. The approach accomplishes the detection of the changed and unchanged areas by designing a deep neural network. The main guideline is to produce a change detection map directly from two images with the trained deep neural network. The method can omit the process of generating a difference image (DI) that shows difference degrees between multitemporal synthetic aperture radar images. Thus, it can avoid the effect of the DI on the change detection results. The learning algorithm for deep architectures includes unsupervised feature learning and supervised fine-tuning to complete classification. The unsupervised feature learning aims at learning the representation of the relationships between the two images. In addition, the supervised fine-tuning aims at learning the concepts of the changed and unchanged pixels. Experiments on real data sets and theoretical analysis indicate the advantages, feasibility, and potential of the proposed method. Moreover, based on the results achieved by various traditional algorithms, respectively, deep learning can further improve the detection performance.

Deep learning applications and challenges in big data analytics Big Data Analytics and Deep Learning are two high-focus of data science. Big Data has become important as many organizations both public and private have been collecting massive amounts of domain-specific information, which can contain useful information about problems such as national intelligence, cyber security, fraud detection, marketing, and medical informatics. Companies such as Google and Microsoft are analyzing large volumes of data for business analysis and decisions, impacting existing and future technology. Deep Learning algorithms extract high-level, complex abstractions as data representations through a hierarchical learning process. Complex abstractions are learnt at a given level based on relatively simpler abstractions formulated in the preceding level in the hierarchy. A key benefit of Deep Learning is the analysis and learning of massive amounts of unsupervised data, making it a valuable tool for Big Data Analytics where raw data is largely unlabeled and un-categorized. In the present study, we explore how Deep Learning can be utilized for addressing some important problems in Big Data Analytics, including extracting complex patterns from massive volumes of data, semantic indexing, data tagging, fast information retrieval, and simplifying discriminative tasks. We also investigate some aspects of Deep Learning research that need further exploration to incorporate specific challenges introduced by Big Data Analytics, including streaming data, high-dimensional data, scalability of models, and distributed computing. We conclude by presenting insights into relevant future works by posing some questions, including defining data sampling criteria, domain adaptation modeling, defining criteria for obtaining useful data abstractions, improving semantic indexing, semi-supervised learning, and active learning.

Revealing Fundamental Physics From The Daya Bay Neutrino Experiment Using Deep Neural Networks Experiments in particle physics produce enormous quantities of data that must be analyzed and interpreted by teams of physicists. This analysis is often exploratory, where scientists are unable to enumerate the possible types of signal prior to performing the experiment. Thus, tools for summarizing, clustering, visualizing and classifying high-dimensional data are essential. In this work, we show that meaningful physical content can be revealed by transforming the raw data into a learned high-level representation using deep neural networks, with measurements taken at the Daya Bay Neutrino Experiment as a case study. We further show how convolutional deep neural networks can provide an effective classification filter with greater than 97% accuracy across different classes of physics events, significantly better than other machine learning approaches.

Efficient Learning of Domain-invariant Image Representations

Empirical Analysis of Predictive Algorithms for Collaborative Filtering Collaborative filtering or recommender systemsuse a database about user preferences topredict additional topics or products a newuser might like. In this paper we describe several algorithms designed for this task, including techniques based on correlation coefficients,vector-based similarity calculations,and statistical Bayesian methods. We comparethe predictive accuracy of the various methods in a set of representative problemdomains. We use two basic classes of evaluation...

Monocular Pedestrian Detection: Survey and Experiments Pedestrian detection is a rapidly evolving area in computer vision with key applications in intelligent vehicles, surveillance, and advanced robotics. The objective of this paper is to provide an overview of the current state of the art from both methodological and experimental perspectives. The first part of the paper consists of a survey. We cover the main components of a pedestrian detection system and the underlying models. The second (and larger) part of the paper contains a corresponding experimental study. We consider a diverse set of state-of-the-art systems: wavelet-based AdaBoost cascade [74], HOG/linSVM [11], NN/LRF [75], and combined shape-texture detection [23]. Experiments are performed on an extensive data set captured onboard a vehicle driving through urban environment. The data set includes many thousands of training samples as well as a 27-minute test sequence involving more than 20,000 images with annotated pedestrian locations. We consider a generic evaluation setting and one specific to pedestrian detection onboard a vehicle. Results indicate a clear advantage of HOG/linSVM at higher image resolutions and lower processing speeds, and a superiority of the wavelet-based AdaBoost cascade approach at lower image resolutions and (near) real-time processing speeds. The data set (8.5 GB) is made public for benchmarking purposes.

On Discriminative vs. Generative Classifiers: A comparison of logistic regression and naive Bayes We compare discriminative and generative learning as typified by logistic regression and naive Bayes. We show, contrary to a widely-held belief that discriminative classifiers are almost always to be preferred, that there can often be two distinct regimes of performance as the training set size is increased, one in which each algorithm does better. This stems from the observation-which is borne out in repeated experiments-that while discriminative learning has lower asymptotic error, a generative classifier may also approach its (higher) asymptotic error much faster.

Formalizing sensing actions—a transition function based approach In presence of incomplete information about the world we need to distinguish between the state of the world and the state of the agent's knowledge about the world. In such a case the agent may need to have at its disposal sensing actions that change its state of knowledge about the world and may need to construct more general plans consisting of sensing actions and conditional statements to achieve its goal. In this paper we first develop a high-level action description language that allows specification of sensing actions and their effects in its domain description and allows queries with conditional plans. We give provably correct translations of domain description in our language to axioms in first-order logic, and relate our formulation to several earlier formulations in the literature. We then analyze the state space of our formulation and develop several sound approximations that have much smaller state spaces. Finally we define regression of knowledge formulas over conditional plans. © 2001 Elsevier Science B.V. All rights reserved.

Weak, strong, and strong cyclic planning via symbolic model checking Planning in nondeterministic domains yields both conceptual and practical difficulties. From the conceptual point of view, different notions of planning problems can be devised: for instance, a plan might either guarantee goal achievement, or just have some chances of success. From the practical point of view, the problem is to devise algorithms that can effectively deal with large state spaces. In this paper, we tackle planning in nondeterministic domains by addressing conceptual and practical problems. We formally characterize different planning problems, where solutions have a chance of success ("weak planning"), are guaranteed to achieve the goal ("strong planning"), or achieve; the goal with iterative trial-and-error strategies ("strong cyclic planning"). In strong cyclic planning, all the executions associated with the solution plan always have a possibility of terminating and, when they do, they are guaranteed to achieve the goal. We present planning algorithms for these problem classes, and prove that they are correct and complete. We implement the algorithms in the MBP planner by using symbolic model checking techniques. We show that our approach is practical with an extensive experimental evaluation: MBP compares positively with state-of-the-art planners, both in terms of expressiveness and in terms of performance.

Iterative Majorization Approach to the Distance-based Discriminant Analysis This paper proposes a method of finding a discriminative linear transformation that enhances the data's degree of conformance to the compactness hypothesis and its inverse. The problem formulation relies on inter-observation distances only, which is shown to improve non-parametric and non-linear classifier performance on benchmark and real-world data sets. The proposed approach is suitable for both binary and multiple-category classification problems, and can be applied as a dimensionality reduction technique. In the latter case, the number of necessary discriminative dimensions can be determined exactly. The sought transformation is found as a solution to an optimization problem using iterative majorization.

GPU-accelerated exhaustive search for third-order epistatic interactions in case-control studies. Interest in discovering combinations of genetic markers from case-control studies, such as Genome Wide Association Studies (GWAS), that are strongly associated to diseases has increased in recent years. Detecting epistasis, i.e. interactions among k markers (k >= 2), is an important but time consuming operation since statistical computations have to be performed for each k-tuple of measured markers. Efficient exhaustive methods have been proposed for k = 2, but exhaustive third-order analyses are thought to be impractical due to the cubic number of triples to be computed. Thus, most previous approaches apply heuristics to accelerate the analysis by discarding certain triples in advance. Unfortunately, these tools can fail to detect interesting interactions. We present GPU3SNP, a fast GPU-accelerated tool to exhaustively search for interactions among all marker-triples of a given case-control dataset. Our tool is able to analyze an input dataset with tens of thousands of markers in reasonable time thanks to two efficient CUDA kernels and efficient workload distribution techniques. For instance, a dataset consisting of 50,000 markers measured from 1000 individuals can be analyzed in less than 22h on a single compute node with 4 NVIDIA GTX Titan boards. (C) 2015 Elsevier ay. All rights reserved.

Optimal Planning for Delete-Free Tasks with Incremental LM-Cut.

Automatic Polytime Reductions of NP Problems into a Fragment of STRIPS.

A Practical, Integer-Linear Programming Model for the Delete-Relaxation in Cost-Optimal Planning. We propose a new integer-linear programming model for the delete relaxation in cost-optimal planning. While a naive formulation of the delete relaxation as IP is impractical, our model incorporates landmarks and relevance-based constraints, resulting in an IP that can be used to directly solve the delete relaxation. We show that our IP model outperforms the previous state-of-the-art solver for delete-free problems. We then use LP relaxation of the IP as a heuristics for a forward search planner, and show that our LP-based solver is competitive with the state-of-the-art for cost-optimal planning.

Minimal Landmarks for Optimal Delete-Free Planning.

Optimizing Plans through Analysis of Action Dependencies and Independencies.

How good is almost perfect? Heuristic search using algorithms such as A* and IDA* is the prevalent method for obtaining optimal sequential solutions for classical planning tasks. Theoretical analyses of these classical search algorithms, such as the well-known results of Pohl, Gaschnig and Pearl, suggest that such heuristic search algorithms can obtain better than exponential scaling behaviour, provided that the heuristics are accurate enough. Here, we show that for a number of common planning benchmark domains, including ones that admit optimal solution in polynomial time, general search algorithms such as A* must necessarily explore an exponential number of search nodes even under the optimistic assumption of almost perfect heuristic estimators, whose heuristic error is bounded by a small additive constant. Our results shed some light on the comparatively bad performance of optimal heuristic search approaches in "simple" planning domains such as GRIPPER. They suggest that in many applications, further improvements in run-time require changes to other parts of the search algorithm than the heuristic estimator.

Recent Advances in AI Planning The past five years have seen dramatic advances in planning algorithms, with an emphasis on propositional methods such as Graphplan and compilers that convert planning problems into propositional CNF formulae for solution via systematic or stochastic SAT methods. Related work on the Deep Space One spacecraft control algorithms advances our understanding of interleaved planning and execution. In this survey,we explain the latest techniques and suggest areas for future research.

Planning with Incomplete Information as Heuristic Search in Belief Space The formulation of planning as heuristic search withheuristics derived from problem representations hasturned out to be a fruitful approach for classical planning.In this paper, we pursue a similar idea in thecontext planning with incomplete information. Planningwith incomplete information can be formulated asa problem of search in belief space, where belief statescan be either sets of states or more generally probabilitydistribution over states. While the formulation (as the...

Probabilistic Planning with Information Gathering and Contingent Execution Most AI representations and algorithms for plan generationhave not included the concept of informationproducingactions (also called diagnostics, or tests,in the decision making literature). We present aplanning representation and algorithm that modelsinformation-producing actions and constructs plansthat exploit the information produced by those actions.We extend the buridan (Kushmerick et al.1994) probabilistic planning algorithm, adapting theaction representation to model the...

PP is closed under intersection In this seminal paper on probabilistic Turing machines, Gill asked whether the class PP is closed under intersection and union. We give a positive answer to this question. We also show that PP is closed under a variety of polynomial-time truth-table reductions. Consequences in complexity theory include the definite collapse and (assuming P ≠ PP) separation of certain query hierarchies over PP. Similar techniques allow us to combine several threshold gates into a single threshold gate. Consequences in the study of circuits include the simulation of circuits with a small number of threshold gates by circuits having only a single threshold gate at the root (perceptrons) and a lower bound on the number of threshold gates that are needed to compute the parity function.

EXPLODE: a lightweight, general system for finding serious storage system errors Storage systems such as file systems, databases, and RAID systems have a simple, basic contract: you give them data, they do not lose or corrupt it. Often they store the only copy, making its irrevocable loss almost arbitrarily bad. Unfortunately, their code is exceptionally hard to get right, since it must correctly recover from any crash at any program point, no matter how their state was smeared across volatile and persistent memory. This paper describes EXPLODE, a system that makes it easy to systematically check real storage systems for errors. It takes user-written, potentially system-specific checkers and uses them to drive a storage system into tricky corner cases, including crash recovery errors. EXPLODE uses a novel adaptation of ideas from model checking, a comprehensive, heavy-weight formal verification technique, that makes its checking more systematic (and hopefully more effective) than a pure testing approach while being just as lightweight. EXPLODE is effective. It found serious bugs in a broad range of real storage systems (without requiring source code): three version control systems, Berkeley DB, an NFS implementation, ten file systems, a RAID system, and the popular VMware GSX virtual machine. We found bugs in every system we checked, 36 bugs in total, typically with little effort.

Learning read-once formulas using membership queries No abstract available.

The Computational Complexity of Agent Design Problems This paper investigates the computational complexity of a fundamental problem in multi-agent systems: given an environment together with a specification of some task, can we construct an agent that will successfully achieve the task in the environment? We refer to this problem as agent design. Using an abstract formal model of agents and their environments, we begin by investigating various possible ways of specifying tasks for agents, and identify two important classes of such tasks. Achievement tasks are those in which an agent is required to bring about one of a specified set of goal states, and maintenance tasks are those in which an agent is required to avoid some specified set of states. We prove that in the most general case the agent design problem is PSPACE-complete for both achievement and maintenance tasks. We briefly discuss the automatic synthesis of agents from task environment specifications, and conclude by discussing related work and presenting some conclusions.

Learning Topic Representation For Smt With Neural Networks Statistical Machine Translation (SMT) usually utilizes contextual information to disambiguate translation candidates. However, it is often limited to contexts within sentence boundaries, hence broader topical information cannot be leveraged. In this paper, we propose a novel approach to learning topic representation for parallel data using a neural network architecture, where abundant topical contexts are embedded via topic relevant monolingual data. By associating each translation rule with the topic representation, topic relevant rules are selected according to the distributional similarity with the source text during SMT decoding. Experimental results show that our method significantly improves translation accuracy in the NIST Chinese-to-English translation task compared to a state-of-the-art baseline.

The dynamic granularity memory system Chip multiprocessors enable continued performance scaling with increasingly many cores per chip. As the throughput of computation outpaces available memory bandwidth, however, the system bottleneck will shift to main memory. We present a memory system, the dynamic granularity memory system (DGMS), which avoids unnecessary data transfers, saves power, and improves system performance by dynamically changing between fine and coarse-grained memory accesses. DGMS predicts memory access granularities dynamically in hardware, and does not require software or OS support. The dynamic operation of DGMS gives it superior ease of implementation and power efficiency relative to prior multi-granularity memory systems, while maintaining comparable levels of system performance.

A logic for default reasoning The need to make default assumptions is frequently encountered in reasoning'about incompletely specified worlds. Inferences sanctioned by default are best viewed as beliefs which may well be modified or rejected by subsequent observations. It is this property which leads to the non.monotonJcity of any logic of defaults. In this paper we propose a logic for default reasoning. We then specialize our treatment to a very large class of commonly occurring defaults. For this class we develop a complete proof theory and show how to interface it with a top down resolution theorem prover. Finally, we provide criteria under which the revision of derived beliefs must be effected.

Extended stable semantics for normal and disjunctive programs

The Stable Model Semantics for Logic Programming We propose a new declarative semantices for logic programs with negation.Its formulation is quite simple;at the same time, it is more general than the iterated fixed point semantics for stratified programs,and is applicable to some useful programs that are not stratified.

Classical Negation in Logic Programs and Disjunctive Databases An important limitation of traditional logic programming as a knowledge representation tool, in comparison with classical logic, is that logic programming does not allow us to deal directly with incomplete information. In order to overcome this limitation, we extend the class of general logic programs by including classical negation, in addition to negation-as-failure. The semantics of such extended programs is based on the method of stable models. The concept of a disjunctive database can be extended in a similar way. We show that some facts of commonsense knowledge can be represented by logic programs and disjunctive databases more easily when classical negation is available.Computationally, classical negation can be eliminated from extended programs by a simple preprocessor. Extended programs are identical to a special case of default theories in the sense of Reiter.

Improvements to the Evaluation of Quantified Boolean Formulae We present a theorem-prover for quantified Boolean formulae and evaluate it on random quantified formulae and formulae that represent problems from automated planning. Even though the notion of quantified Boolean formula is theoretically important, automated reasoning with QBF has not been thoroughly investigated. Universal quantifiers are needed in representing many computational problems that cannot be easily translated to the propositional logic and solved by satisfiability algorithms. Therefore efficient reasoning with QBF is important. The Davis-Putnam procedure can be extended to evaluate quantified Boolean formulae. A straightforward algorithm of this kind is not very efficient. We identify universal quantifiers as the main area where improvements to the basic algorithm can be made. We present a number of techniques for reducing the amount of search that is needed, and evaluate their effectiveness by running the algorithm on a collection of formulae obtained from planning and generated randomly. For the structured problems we consider, the techniques lead to a dramatic speed-up.

Multi-level transaction management for complex objects: implementation, performance, parallelism Multi-level transactions are a variant of open-nested transactions in which the subtransactions correspond to operations at different levels of a layered system architecture. They allow the exploitation of semantics of high-level operations to increase concurrency. As a consequence, undoing a transaction requires compensation of completed subtransactions. In addition, multi-level recovery methods must take into consideration that high-level operations are not necessarily atomic if multiple pages are updated in a single subtransaction. This article presents algorithms for multi-level transaction management that are implemented in the database kernel system (DASDBS). In particular, we show that multi-level recovery can be implemented in an efficient way. We discuss performance measurements using a synthetic benchmark for processing complex objects in a multi-user environment. We show that multi-level transaction management can be extended easily to cope with parallel subtransactions within a single transaction. Performance results are presented with varying degrees of inter- and intratransaction parallelism.

Nonlinear component analysis as a kernel eigenvalue problem A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in high-dimensional feature spaces, related to input space by some nonlinear map-for instance, the space of all possible five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

Performance of a mirrored disk in a real-time transaction system Disk mirroring has found widespread use in computer systems as a method for providing fault tolerance. In addition to increasing reliability, a mirrored disk can also reduce I/O response time by supporting the execution of parallel I/O requests. The improvement in I/O efficiency is extremely important in a real-time system, where each computational entity carries a deadline. In this paper, we present two classes of real-time disk scheduling policies, RT-DMQ and RT-CMQ, for a mirrored disk I/O subsystem and examine their performance in an integrated real-time transaction system. The real-time transaction system model is validated on a real-time database testbed, called RT-CARAT. The performance results show that a mirrored disk I/O subsystem can decrease the fraction of transactions that miss their deadlines over a single disk system by 68%. Our results also reveal the importance of real-time scheduling policies, which can lead up to a 17% performance improvement over non-real-time policies in terms of minimizing the transaction loss ratio.

Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing Solving the SLAM (simultaneous localization and mapping) problem is one way to enable a robot to explore, map, and navigate in a previously unknown environment. Smoothing approaches have been investigated as a viable alternative to extended Kalman filter (EKF)-based solutions to the problem. In particular, approaches have been looked at that factorize either the associated information matrix or the measurement Jacobian into square root form. Such techniques have several significant advantages over the EKF: they are faster yet exact; they can be used in either batch or incremental mode; are better equipped to deal with non-linear process and measurement models; and yield the entire robot trajectory, at lower cost for a large class of SLAM problems. In addition, in an indirect but dramatic way, column ordering heuristics automatically exploit the locality inherent in the geographic nature of the SLAM problem. This paper presents the theory underlying these methods, along with an interpretation of factorization in terms of the graphical model associated with the SLAM problem. Both simulation results and actual SLAM experiments in large-scale environments are presented that underscore the potential of these methods as an alternative to EKF-based approaches.

An A Prolog decision support system for the Space Shuttle The goal of this paper is to test if a programming methodology based on the declarative language A-Prolog and the systems for computing answer sets of such programs, can be successfully applied to the development of medium size knowledge-intensive applications. We report on a successful design and development of such a system controlling some of the functions of the Space Shuttle.

Domain adaptation for object recognition: An unsupervised approach Adapting the classifier trained on a source domain to recognize instances from a new target domain is an important problem that is receiving recent attention. In this paper, we present one of the first studies on unsupervised domain adaptation in the context of object recognition, where we have labeled data only from the source domain (and therefore do not have correspondences between object categories across domains). Motivated by incremental learning, we create intermediate representations of data between the two domains by viewing the generative subspaces (of same dimension) created from these domains as points on the Grassmann manifold, and sampling points along the geodesic between them to obtain subspaces that provide a meaningful description of the underlying domain shift. We then obtain the projections of labeled source domain data onto these subspaces, from which a discriminative classifier is learnt to classify projected data from the target domain. We discuss extensions of our approach for semi-supervised adaptation, and for cases with multiple source and target domains, and report competitive results on standard datasets.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Reasoning about actions: non-deterministic effects, constraints, and qualification In this paper we propose the language of 'state specifications' to uniformly specify effect of actions, executability condition of actions, and dynamic and static constraints. This language allow? us to be able to express effects of action and constraints with same first order representation but different intuitive behavior to be specified differently. We then discuss how we can use state specifications to extend the action description languages A and Lo.

Logic programming and knowledge representation-the A-prolog perspective In this paper we give a short introduction to logic programming approach to knowledge representation and reasoning. The intention is to help the reader to develop a `feel' for the field's history and some of its recent developments. The discussion is mainly limited to logic programs under the answer set semantics. For understanding of approaches to logic programming built on well-founded semantics, general theories of argumentation, abductive reasoning, etc., the reader is referred to other publications.

Expressive Reasoning about Action in Nondeterministic Polynomial Time The rapid development of efficient heuristics for deciding satisfiability for propositional logic motivates thorough investigations of the usability of NP-complete problems in general. In this paper we introduce a logic of action and change which is expressive in the sense that it can represent most propositional benchmark examples in the literature, and some new examples involving parallel composition of actions, and actions that may or may not be executed. We prove that satisfiability of a scenario in this logic is NP-complete, and that it subsumes an NP-complete logic (which in turn includes a nontrivial polynomial-time fragment) previously introduced by Drakengren and Bjareland.

Effect of knowledge representation on model based planning: experiments using logic programming encodings

Language independence and language tolerance in logic programs The consequences of a logic program depend in general upon both the rulesof the program and its language. However the consequences of some programsare independent of the choice of language, while others depend on thelanguage of the program in only a restricted way. In this paper, we definenotions of language independence and language tolerance corresponding tothese two cases. Furthermore, we show that there are syntactically-definedclasses of programs that are language independent and...

Efficient Temporal Reasoning In The Cached Event Calculus This article deals with the problem of providing Kowalski and Sergot's event calculus, extended with context dependency, with an efficient implementation in a logic programming framework. Despite a widespread recognition that a positive solution to efficiency issues is necessary to guarantee the computational feasibility of existing approaches to temporal reasoning, the problem of analyzing the complexity of temporal reasoning programs has been largely overlooked. This article provides a mathematical analysis of the efficiency of query and update processing in the event calculus and defines a cached version of the calculus that (i) moves computational complexity from query to update processing and (ii) features an absolute improvement of performance, because query processing in the event calculus costs much more than update processing in the proposed cached version.

Action Languages Action languages are formal models of parts of the natural languagethat are used for talking about the effects of actions. This article is acollection of definitions related to action languages that may be usefulas a reference in future publications.1 IntroductionThis article is a collection of definitions related to action languages. Itdoes not provide a comprehensive discussion of the subject, and does notcontain a complete bibliography, but it may be useful as a reference in...

Ramification and causality The ramification problem in the context of commonsense reasoning about actions andchange names the challenge to accommodate actions whose execution causes indirecteffects. Not being part of the respective action specification, such effects are consequencesof general laws describing dependencies between components of the world description. Wepresent a general approach to this problem which incorporates causality, formalized bydirected relations between two single effects stating that, under ...

The FF planning system: fast plan generation through heuristic search We describe and evaluate the algorithmic techniques that are used in the FF planning system. Like the HSP system, FF relies on forward state space search, using a heuristic that estimates goal distances by ignoring delete lists. Unlike HSP's heuristic, our method does not assume facts to be independent. We introduce a novel search strategy that combines hill-climbing with systematic search, and we show how other powerful heuristic information can be extracted and used to prune the search space. FF was the most successful automatic planner at the recent AIPS-2000 planning competition. We review the results of the competition, give data for other benchmark domains, and investigate the reasons for the runtime performance of FF compared to HSP.

Signed data dependencies in logic programs Logic programming with negation has been given a declarative semantics by Clark's completed database (CDB), and one can consider the consequences of the CDB in either two-valued or three-valued logic. Logic programming also has a proof theory given by SLDNF derivations. Assuming the data-dependency condition of strictness , we prove that the two-valued and three-valued semantics are equivalent. Assuming allowedness (a condition on occurrences of variables), we prove that SLDNF is complete for the three-valued semantics. Putting these two results together, we have completeness of SLDNF deductions for strict and allowed databases and queries under the standard two-valued semantics. This improves a theorem of Cavedon and Lloyd, who obtained the same result under the additional assumption of stratifiability .

Representing paraconsistent reasoning via quantified propositional logic Quantified propositional logic is an extension of classical propositional logic where quantifications over atomic formulas are permitted. As such, quantified propositional logic is a fragment of second-order logic, and its sentences are usually referred to as quantified Boolean formulas (QBFs). The motivation to study quantified propositional logic for paraconsistent reasoning is based on two fundamental observations. Firstly, in recent years, practicably efficient solvers for quantified propositional logic have been presented. Secondly, complexity results imply that there is a wide range of paraconsistent reasoning problems which can be efficiently represented in terms of QBFs. Hence, solvers for QBFs can be used as a core engine in systems prototypically implementing several of such reasoning tasks, most of them lacking concrete realisations. To this end, we show how certain paraconsistent reasoning principles can be naturally formulated or reformulated by means of quantified Boolean formulas. More precisely, we describe polynomial-time constructible encodings providing axiomatisations of the given reasoning tasks. In this way, a whole variety of a priori distinct approaches to paraconsistent reasoning become comparable in a uniform setting.

Distributed operating systems Distributed operating systems have many aspects in common with centralized ones, but they also differ in certain ways. This paper is intended as an introduction to distributed operating systems, and especially to current university research about them. After a discussion of what constitutes a distributed operating system and how it is distinguished from a computer network, various key design issues are discussed. Then several examples of current research projects are examined in some detail, namely, the Cambridge Distributed Computing System, Amoeba, V, and Eden.

A polynomial time solution for protein chain pair simplification under the discrete fréchet distance The comparison and simplification of polygonal chains is an important and active topic in many areas of research. In the study of protein structure alignment and comparison, a lot of work has been done using RMSD as the distance measure. This method has certain drawbacks, and thus recently, the discrete Fréchet distance was applied to the problem of protein (backbone) structure alignment and comparison with promising results. Another important area within protein structure research is visualization, due to the number of nodes along each backbone. Protein chain backbones can have as many as 500~600 α -carbon atoms which constitute the vertices in the comparison. Even with an excellent alignment, the similarity of two polygonal chains can be very difficult to see visually unless the two chains are nearly identical. To address this issue, the chain pair simplification problem (CPS-3F) was proposed in 2008 to simultaneously simplify both chains with respect to each other under the discrete Fréchet distance. It is unknown whether CPS-3F is NP -complete, and so heuristic methods have been developed. Here, we first define a version of CPS-3F, denoted CPS-3F+, and prove that it is polynomially solvable by presenting a dynamic programming solution. Then we compare the CPS-3F+ solutions with previous empirical results, and further demonstrate some of the benefits of the simplified comparison. Finally, we discuss future work and implications along with a web-based software implementation, named FPACT (The Fréchet-based Protein Alignment & Comparison Tool), allowing users to align, simplify, and compare protein backbone chains using methods based on the discrete Fréchet distance.

Exploring Sequence Alignment Algorithms On Fpga-Based Heterogeneous Architectures With the rapid development of DNA sequencer, the rate of data generation is rapidly outpacing the rate at which it can be computationally processed. Traditional sequence alignment based on PC cannot fulfill the increasing demand. Accelerating the algorithm using FPGA provides the better performance compared to the other platforms. This paper will explain and classify the current sequence alignment algorithms. In addition, we analyze the different types of sequence alignment algorithms and present the taxonomy of FPGA-based sequence alignment implementations. This work will conclude the current solutions and provide a reference to further accelerating sequence alignment on a FPGA-based heterogeneous architecture.

Chirp: a practical global filesystem for cluster and Grid computing Traditional distributed £lesystem technologies designed for local and campus area networks do not adapt well to wide area grid computing environments. To address this problem, we have designed the Chirp distributed £lesystem, which is designed from the ground up to meet the needs of grid computing. Chirp is easily deployed without special privileges, provides strong andexible security mechanisms, tunable consistency semantics, and clustering to increase capacity and throughput. We demonstrate that many of these features also provide order- of-magnitude performance increases over wide area networks. We describe three applications in bioinformatics, biometrics, and gamma ray physics that each em- ploy Chirp to attack large scale data intensive problems.

/scratch as a cache: rethinking HPC center scratch storage To sustain emerging data-intensive scientific applications, High Performance Computing (HPC) centers invest a notable fraction of their operating budget on a specialized fast storage system, scratch space, which is designed for storing the data of currently running and soon-to-run HPC jobs. Instead, it is often used as a standard file system, wherein users arbitrarily store their data, without any consideration to the center's overall performance. To remedy this, centers periodically scan the scratch in an attempt to purge transient and stale data.This practice of supporting a cache workload using a file system and disjoint tools for staging and purging results in suboptimal use of the scratch space. In this paper, we address the above issues by proposing a new perspective, where the HPC scratch space is treated as a cache, and data population, retention, and eviction tools are integrated with scratch management. In our approach, data is moved to the scratch space only when it is needed, and unneeded data is removed as soon as possible. We also design a new job-workflow-aware caching policy that leverages user-supplied hints for managing the cache. Our evaluation using three-year job logs from the Jaguar supercomputer, shows that compared to the widely-used purge approach, workflow-aware caching optimizes scratch utilization by reducing the average amount of data read by 9.3%, and by reducing job scheduling delays associated with data staging, on average, by 282.0%.

Minerva: An automated resource provisioning tool for large-scale storage systems Enterprise-scale storage systems, which can contain hundreds of host computers and storage devices and up to tens of thousands of disks and logical volumes, are difficult to design. The volume of choices that need to be made is massive, and many choices have unforeseen interactions. Storage system design is tedious and complicated to do by hand, usually leading to solutions that are grossly over-provisioned, substantially under-performing or, in the worst case, both.To solve the configuration nightmare, we present minerva: a suite of tools for designing storage systems automatically. Minerva uses declarative specifications of application requirements and device capabilities; constraint-based formulations of the various sub-problems; and optimization techniques to explore the search space of possible solutions.This paper also explores and evaluates the design decisions that went into Minerva, using specialized micro- and macro-benchmarks. We show that Minerva can successfully handle a workload with substantial complexity (a decision-support database benchmark). Minerva created a 16-disk design in only a few minutes that achieved the same performance as a 30-disk system manually designed by human experts. Of equal importance, Minerva was able to predict the resulting system's performance before it was built.

Track-Aligned Extents: Matching Access Patterns to Disk Drive Characteristics Track-aligned extents (traxtents) utilize disk-specific knowledge to match access patterns to the strengths of modern disks. By allocating and accessing related data on disk track boundaries, a system can avoid most rotational latency and track crossing overheads. Avoiding these overheads can increase disk access efficiency by up to 50% for mid-sized requests (100-500KB). This paper describes traxtents, algorithms for detecting track boundaries, and some uses of traxtents in file systems and video servers. For large-file workloads, a version of FreeBSD's FFS implementation that exploits traxtents reduces application run times by up to 20% compared to the original version. A video server using traxtent-based requests can support 56% more concurrent streams at the same startup latency and buffer space. For LFS, 44% lower overall write cost for track-sized segments can be achieved.

Logic programs with classical negation

Logic programming and knowledge representation In this paper, we review recent work aimed at the application of declarative logic programming to knowledge representation in artificial intelligence. We consider extensions of the language of definite logic programs by classical (strong) negation, disjunction, and some modal operators and show how each of the added features extends the representational power of the language.

The contract net protocol: high-level communication and control in a distributed problem solver The contract net protocol has been developed to specify problem-solving communication and control for nodes in a distributed problem solver. Task distribution is affected by a negotiation process, a discussion carried on between nodes with tasks to be executed and nodes that may be able to execute those tasks.

A trace-driven analysis of the UNIX 4.2 BSD file system

Proceedings of the Sixteenth International Joint Conference on Artificial Intelligence, IJCAI 99, Stockholm, Sweden, July 31 - August 6, 1999. 2 Volumes, 1450 pages

Fine-Grained Mobility in the Emerald System (Extended Abstract)

Normal forms for answer sets programming Normal forms for logic programs under stable/answer set semantics are introduced. We argue that these forms can simplify the study of program properties, mainly consistency. The first normal form, called the kernel of the program, is useful for studying existence and number of answer sets. A kernel program is composed of the atoms which are undefined in the Well-founded semantics, which are those that directly affect the existence of answer sets. The body of rules is composed of negative literals only. Thus, the kernel form tends to be significantly more compact than other formulations. Also, it is possible to check consistency of kernel programs in terms of colorings of the Extended Dependency Graph program representation which we previously developed. The second normal form is called 3-kernel. A 3-kernel program is composed of the atoms which are undefined in the Well-founded semantics. Rules in 3-kernel programs have at most two conditions, and each rule either belongs to a cycle, or defines a connection between cycles. 3-kernel programs may have positive conditions. The 3-kernel normal form is very useful for the static analysis of program consistency, i.e. the syntactic characterization of existence of answer sets. This result can be obtained thanks to a novel graph-like representation of programs, called Cycle Graph which presented in the companion article Costantini (2004b).

ARIMA time series modeling and forecasting for adaptive I/O prefetching Bursty application I/O patterns, together with transfer limited storage devices, combine to create a major I/O bottleneck on parallel systems. This paper explores the use of time series models to forecast application I/O request times, then prefetching I/O requests during computation intervals to hide I/O latency. Experimental results with I/O intensive scientific codes show performance improvements compared to standard UNIX prefetching strategies.

An Unsupervised Feature Learning Approach to Improve Automatic Incident Detection. Sophisticated automatic incident detection (AID) technology plays a key role in contemporary transportation systems. Though many papers were devoted to study incident classification algorithms, few study investigated how to enhance feature representation of incidents to improve AID performance. In this paper, we propose to use an unsupervised feature learning algorithm to generate higher level features to represent incidents. We used real incident data in the experiments and found that effective feature mapping function can be learnt from the data crosses the test sites. With the enhanced features, detection rate (DR), false alarm rate (FAR) and mean time to detect (MTTD) are significantly improved in all of the three representative cases. This approach also provides an alternative way to reduce the amount of labeled data, which is expensive to obtain, required in training better incident classifiers since the feature learning is unsupervised. © 2012 IEEE.

Protecting RAID Arrays against Unexpectedly High Disk Failure Rates Disk failure rates vary so widely among different makes and models that designing storage solutions for the worst case scenario is a losing proposition. The approach we propose here is to design our storage solutions for the most probable case while incorporating in our design the option of adding extra redundancy when we find out that its disks are less reliable than expected. To illustrate our proposal, we show how to increase the reliability of existing two-dimensional disk arrays with n^2 data elements and 2n parity elements by adding n additional parity elements that will mirror the contents of half the existing parity elements. Our approach offers the three advantages of being easy to deploy, not affecting the complexity of parity calculations, and providing a five-year reliability of 99.999 percent in the face of catastrophic levels of data loss where the array would lose up to a quarter of its storage capacity in a year.

Slow feature analysis: unsupervised learning of invariances. Invariant features of temporally varying signals are useful for analysis and classification. Slow feature analysis (SFA) is a new method for learning invariant or slowly varying features from a vectorial input signal. It is based on a nonlinear expansion of the input signal and application of principal component analysis to this expanded signal and its time derivative. It is guaranteed to find the optimal solution within a family of functions directly and can learn to extract a large number of decorrelated features, which are ordered by their degree of invariance. SFA can be applied hierarchically to process high-dimensional input signals and extract complex features. SFA is applied first to complex cell tuning properties based on simple cell output, including disparity and motion. Then more complicated input-output functions are learned by repeated application of SFA. Finally, a hierarchical network of SFA modules is presented as a simple model of the visual system. The same unstructured network can learn translation, size, rotation, contrast, or, to a lesser degree, illumination invariance for one-dimensional objects, depending on only the training stimulus. Surprisingly, only a few training objects suffice to achieve good generalization to new objects. The generated representation is suitable for object recognition. Performance degrades if the network is trained to learn multiple invariances simultaneously.

Rational Kernels: Theory and Algorithms Many classification algorithms were originally designed for fixed-size vectors. Recent applications in text and speech processing and computational biology require however the analysis of variable-length sequences and more generally weighted automata. An approach widely used in statistical learning techniques such as Support Vector Machines (SVMs) is that of kernel methods, due to their computational efficiency in high-dimensional feature spaces. We introduce a general family of kernels based on weighted transducers or rational relations, rational kernels , that extend kernel methods to the analysis of variable-length sequences or more generally weighted automata. We show that rational kernels can be computed efficiently using a general algorithm of composition of weighted transducers and a general single-source shortest-distance algorithm. Not all rational kernels are positive definite and symmetric (PDS), or equivalently verify the Mercer condition, a condition that guarantees the convergence of training for discriminant classification algorithms such as SVMs. We present several theoretical results related to PDS rational kernels. We show that under some general conditions these kernels are closed under sum, product, or Kleene-closure and give a general method for constructing a PDS rational kernel from an arbitrary transducer defined on some non-idempotent semirings. We give the proof of several characterization results that can be used to guide the design of PDS rational kernels. We also show that some commonly used string kernels or similarity measures such as the edit-distance, the convolution kernels of Haussler, and some string kernels used in the context of computational biology are specific instances of rational kernels. Our results include the proof that the edit-distance over a non-trivial alphabet is not negative definite, which, to the best of our knowledge, was never stated or proved before. Rational kernels can be combined with SVMs to form efficient and powerful techniques for a variety of classification tasks in text and speech processing, or computational biology. We describe examples of general families of PDS rational kernels that are useful in many of these applications and report the result of our experiments illustrating the use of rational kernels in several difficult large-vocabulary spoken-dialog classification tasks based on deployed spoken-dialog systems. Our results show that rational kernels are easy to design and implement and lead to substantial improvements of the classification accuracy.

An Information Measure For Classification

Self Supervised Boosting Boosting algorithms and successful applications thereof abound for clas- sification and regression learning problems, but not for unsupervised learning. We propose a sequential approach to adding features to a ran- dom field model by training them to improve classification performance between the data and an equal-sized sample of "negative examples" gen- erated from the model's current estimate of the data density. Training in each boosting round proceeds in three stages: first we sample negative examples from the model's current Boltzmann distribution. Next, a fea- ture is trained to improve classification performance between data and negative examples. Finally, a coefficient is learned which determines the importance of this feature relative to ones already in the pool. Negative examples only need to be generated once to learn each new feature. The validity of the approach is demonstrated on binary digits and continuous synthetic data.

A two-layer ICA-like model estimated by score matching Capturing regularities in high-dimensional data is an important problem in machine learning and signal processing. Here we present a statistical model that learns a nonlinear representation from the data that reflects abstract, invariant properties of the signal without making requirements about the kind of signal that can be processed. The model has a hierarchy of two layers, with the first layer broadly corresponding to Independent Component Analysis (ICA) and a second layer to represent higher order structure. We estimate the model using the mathematical framework of Score Matching (SM), a novel method for the estimation of non-normalized statistical models. The model incorporates a squaring nonlinearity, which we propose to be suitable for forming a higher-order code of invariances. Additionally the squaring can be viewed as modelling subspaces to capture residual dependencies, which linear models cannot capture.

Implicit Density Estimation by Local Moment Matching to Sample from Auto-Encoders Recent work suggests that some auto-encoder variants do a good job of capturing the local manifold structure of the unknown data generating density. This paper contributes to the mathematical understanding of this phenomenon and helps define better justified sampling algorithms for deep learning based on auto-encoder variants. We consider an MCMC where each step samples from a Gaussian whose mean and covariance matrix depend on the previous state, defines through its asymptotic distribution a target density. First, we show that good choices (in the sense of consistency) for these mean and covariance functions are the local expected value and local covariance under that target density. Then we show that an auto-encoder with a contractive penalty captures estimators of these local moments in its reconstruction function and its Jacobian. A contribution of this work is thus a novel alternative to maximum-likelihood density estimation, which we call local moment matching. It also justifies a recently proposed sampling algorithm for the Contractive Auto-Encoder and extends it to the Denoising Auto-Encoder.

A Sparse and Locally Shift Invariant Feature Extractor Applied to Document Images We describe an unsupervised learning algorithm for ex- tracting sparse and locally shift-invariant features. We also devise a principled procedure for learning hierarchies of in- variant features. Each feature detector is composed of a set of trainable convolutional filters followed by a max-pooling layer over non-overlapping windows, and a point-wise sig- moid non-linearity. A second stage of more invariant fea- tures is fed with patches provided by the first stage feature extractor, and is trained in the same way. The method is used to pre-train the first four layers of a deep convolutional network which achieves state-of-the-art performance on the MNIST dataset of handwritten digits. The final testing error rate is equal to 0.42%. Preliminary experiments on com- pression of bitonal document images show very promising results in terms of compression ratio and reconstruction er- ror.

Extracting distributed representations of concepts and relations from positive and negative propositions Linear relational embedding (LRE) was introduced previously by the authors (1999) as a means of extracting a distributed representation of concepts from relational data. The original formulation cannot use negative information and cannot properly handle data in which there are multiple correct answers. In this paper we propose an extended formulation of LRE that solves both these problems. We present results in two simple domains, which show that learning leads to good generalization

A Scalable Hierarchical Distributed Language Model Neural probabilistic language models (NPLMs) have been shown to be competi- tive with and occasionally superior to the widely-used n-gram language models. The main drawback of NPLMs is their extremely long training and testing times. Morin and Bengio have proposed a hierarchical language model built around a binary tree of words, which was two orders of magnitude faster than the non- hierarchical model it was based on. However, it performed considerably worse than its non-hierarchical counterpart in spite of using a wo rd tree created using expert knowledge. We introduce a fast hierarchical language model along with a simple feature-based algorithm for automatic construction of word trees from the data. We then show that the resulting models can outperform non-hierarchical neural models as well as the best n-gram models.

A Better Way to Pretrain Deep Boltzmann Machines.

A deep graph embedding network model for face recognition In this paper, we propose a new deep learning network “GENet”, it combines the multi-layer network architecture and graph embedding framework. Firstly, we use simplest unsupervised learning PCA/LDA as first layer to generate the low-level feature. Secondly, many cascaded dimensionality reduction layers based on graph embedding framework are applied to GENet. Finally, a linear SVM classifier is used to classify dimension-reduced features. The experiments indicate that higher classification accuracy can be obtained by this algorithm on the CMU-PIE, ORL, Extended Yale B dataset.

Two counterexamples related to Baker's approach to the frame problem Andrew Baker's approach to reasoning about actions is the mo st robust circumscriptive approach currently known. Investigation of its applicabil ity to nondeterministic actions reveals that this approach does not allow us to draw some intuitively plausible conclu- sions. Also, it does not always generate the proper existenc e of situations axiom. The limitations are traced to an unexpected interference of the axioms encoding observations with the minimization. A modification that avoids the shortcomings is suggested.

A systematic approach to system state restoration during storage controller micro-recovery Micro-recovery, or failure recovery at a fine granularity, is a promising approach to improve the recovery time of software for modern storage systems. Instead of stalling the whole system during failure recovery, micro-recovery can facilitate recovery by a single thread while the system continues to run. A key challenge in performing micro-recovery is to be able to perform efficient and effective state restoration while accounting for dynamic dependencies between multiple threads in a highly concurrent environment. We present Log(Lock), a practical and flexible architecture for performing state restoration without re-architecting legacy code. We formally model thread dependencies based on accesses to both shared state and resources. The Log(Lock) execution model tracks dependencies at runtime and captures the failure context through the restoration level. We develop restoration protocols based on recovery points and restoration levels that identify when micro-recovery is possible and the recovery actions that need to be performed for a given failure context. We have implemented Log(Lock) in a real enterprise storage controller. Our experimental evaluation shows that Log(Lock)-enabled micro-recovery is efficient. It imposes

Super-Solutions: Succinctly Representing Solutions in Abductive Annotated Probabilistic Temporal Logic Annotated Probabilistic Temporal (APT) logic programs are a form of logic programs that allow users to state (or systems to automatically learn) rules of the form “formula G becomes true Δt time units after formula F became true with &ell; to u&percnt; probability.” In this article, we deal with abductive reasoning in APT logic: given an APT logic program Π, a set of formulas H that can be “added” to Π, and a (temporal) goal g, is there a subset S of H such that Π ∪ S is consistent and entails the goal g&quest; In general, there are many different solutions to the problem and some of them can be highly repetitive, differing only in some unimportant temporal aspects. We propose a compact representation called super-solutions that succinctly represent sets of such solutions. Super-solutions are compact, but lossless representations of sets of such solutions. We study the complexity of existence of basic, super-, and maximal super-solutions as well as check if a set is a solution/super-solution/maximal super-solution. We then leverage a geometric characterization of the problem to suggest a set of pruning strategies and interesting properties that can be leveraged to make the search of basic and super-solutions more efficient. We propose correct sequential algorithms to find solutions and super-solutions. In addition, we develop parallel algorithms to find basic and super-solutions.

Scalable Object Retrieval with Compact Image Representation from Generic Object Regions In content-based visual object retrieval, image representation is one of the fundamental issues in improving retrieval performance. Existing works adopt either local SIFT-like features or holistic features, and may suffer sensitivity to noise or poor discrimination power. In this article, we propose a compact representation for scalable object retrieval from few generic object regions. The regions are identified with a general object detector and are described with a fusion of learning-based features and aggregated SIFT features. Further, we compress feature representation in large-scale image retrieval scenarios. We evaluate the performance of the proposed method on two public ground-truth datasets, with promising results. Experimental results on a million-scale image database demonstrate superior retrieval accuracy with efficiency gain in both computation and memory usage.

Multimedia answering: enriching text QA with media information Existing community question-answering forums usually provide only textual answers. However, for many questions, pure texts cannot provide intuitive information, while image or video contents are more appropriate. In this paper, we introduce a scheme that is able to enrich text answers with image and video information. Our scheme investigates a rich set of techniques including question/answer classification, query generation, image and video search reranking, etc. Given a question and the community-contributed answer, our approach is able to determine which type of media information should be added, and then automatically collects data from Internet to enrich the textual answer. Different from some efforts that attempt to directly answer questions with image and video data, our approach is built based on the community-contributed textual answers and thus it is more feasible and able to deal with more complex questions. We have conducted empirical study on more than 3,000 QA pairs and the results demonstrate the effectiveness of our approach.

Disease Inference from Health-Related Questions via Sparse Deep Learning Automatic disease inference is of importance to bridge the gap between what online health seekers with unusual symptoms need and what busy human doctors with biased expertise can offer. However, accurately and efficiently inferring diseases is non-trivial, especially for community-based health services due to the vocabulary gap, incomplete information, correlated medical concepts, and limited high quality training samples. In this paper, we first report a user study on the information needs of health seekers in terms of questions and then select those that ask for possible diseases of their manifested symptoms for further analytic. We next propose a novel deep learning scheme to infer the possible diseases given the questions of health seekers. The proposed scheme comprises of two key components. The first globally mines the discriminant medical signatures from raw features. The second deems the raw features and their signatures as input nodes in one layer and hidden nodes in the subsequent layer, respectively. Meanwhile, it learns the inter-relations between these two layers via pre-training with pseudolabeled data. Following that, the hidden nodes serve as raw features for the more abstract signature mining. With incremental and alternative repeating of these two components, our scheme builds a sparsely connected deep architecture with three hidden layers. Overall, it well fits specific tasks with fine-tuning. Extensive experiments on a real-world dataset labeled by online doctors show the significant performance gains of our scheme.

Extended stable semantics for normal and disjunctive programs

The nature of statistical learning theory~. First Page of the Article

A machine program for theorem-proving The programming of a proof procedure is discussed in connection with trial runs and possible improvements.

An Introduction to Least Commitment Planning Recent developments have clarified the process of generating partially ordered, partially specified sequences of actions whose execution will achieve an agent's goal. This article summarizes a progression of least commitment planners, starting with one that handles the simple STRIPS representation and ending with UCOPOP a planner that manages actions with disjunctive precondition, conditional effects, and universal quantification over dynamic universes. Along the way, I explain how Chapman's formulation of the modal truth criterion is misleading and why his NP-completeness result for reasoning about plans with conditional effects does not apply to UCOPOP.

Equilibria and steering laws for planar formations This paper presents a Lie group setting for the problem of control of formations, as a natural outcome of the analysis of a planar two-vehicle formation control law. The vehicle trajectories are described using the planar Frenet–Serret equations of motion, which capture the evolution of both the vehicle position and orientation for unit-speed motion subject to curvature (steering) control. The set of all possible (relative) equilibria for arbitrary G-invariant curvature controls is described (where G=SE(2) is a symmetry group for the control law), and a global convergence result for the two-vehicle control law is proved. An n-vehicle generalization of the two-vehicle control law is also presented, and the corresponding (relative) equilibria for the n-vehicle problem are characterized. Work is on-going to discover stability and convergence results for the n-vehicle problem.

Pushing the envelope: planning, propositional logic, and stochastic search Planning is a notoriously hard combinatorial search problem. In many interesting domains, current planning algorithms fail to scale up gracefully. By combining a general, stochastic search algorithm and appropriate problem encodings based on propositional logic, we are able to solve hard planning problems many times faster than the best current planning systems. Although stochastic methods have been shown to be very effective on a wide range of scheduling problems, this is the first demonstration of its power on truly challenging classical planning instances. This work also provides a new perspective on representational issues in planning.

A Completeness Result for SLDNF-Resolution Because of the possibility of floundering and infinite derivations, SLDNF-resolution is, in general, not complete. The classical approach [17] to get a completeness result is to restrict the attention to normal programs P and normal goals G, such that P or {G} is allowed and P is hierarchical. Unfortunately, the class of all normal programs and all normal goals meeting these requirements is not powerful enough to be of great practical importance. But after refining the concept of allowedness by taking modes [12] into account, we can broaden the notion of a hierarchical program, and thereby define a subclass of the class of normal programs and normal goals which is powerful enough to compute all primitive recursive functions without losing the completeness of SLDNF-resolution.

Diagnostic reasoning with A-Prolog In this paper, we suggest an architecture for a software agent which operates a physical device and is capable of making observations and of testing and repairing the device's components. We present simplified definitions of the notions of symptom, candidate diagnosis, and diagnosis which are based on the theory of action language ${\cal AL}$. The definitions allow one to give a simple account of the agent's behavior in which many of the agent's tasks are reduced to computing stable models of logic programs.

iSAM: Incremental Smoothing and Mapping In this paper, we present incremental smoothing and mapping (iSAM), which is a novel approach to the simultaneous localization and mapping problem that is based on fast incremental matrix factorization. iSAM provides an efficient and exact solution by updating a QR factorization of the naturally sparse smoothing information matrix, thereby recalculating only those matrix entries that actually change. iSAM is efficient even for robot trajectories with many loops as it avoids unnecessary fill-in in the factor matrix by periodic variable reordering. Also, to enable data association in real time, we provide efficient algorithms to access the estimation uncertainties of interest based on the factored information matrix. We systematically evaluate the different components of iSAM as well as the overall algorithm using various simulated and real-world datasets for both landmark and pose-only settings.

When Multivariate Forecasting Meets Unsupervised Feature Learning - Towards a Novel Anomaly Detection Framework for Decision Support. Many organizations adopt information technologies to make intelligent decisions during operations. Time-series data plays a crucial role in supporting such decision making processes. Though current studies on time-series based decision making provide reasonably well results, the anomaly detection essence underling most of the scenarios and the plenitude of unlabeled data are largely overlooked and left unexplored. We argue that by using multivariate forecasting and unsupervised feature learning, these two important research gaps could be filled. We carried out two experiments in this study to testify our approach and the results showed that decision support performance was significantly improved. We also proposed a novel framework to integrate the two methods so that our approach may be generalized to a larger problem domain. We discussed the advantages, the limitations and the future work of our study. Both practical and theoretical contributions were also discussed in the paper. © 2012 by the AIS/ICIS Administrative Office All rights reserved.

Learning A Lexical Simplifier Using Wikipedia In this paper we introduce a new lexical simplification approach. We extract over 30K candidate lexical simplifications by identifying aligned words in a sentence-aligned corpus of English Wikipedia with Simple English Wikipedia. To apply these rules, we learn a feature-based ranker using SVMnk trained on a set of labeled simplifications collected using Amazon's Mechanical Turk. Using human simplifications for evaluation, we achieve a precision of 76% with changes in 86% of the examples.

Algebraic multiscale method for flow in heterogeneous porous media with embedded discrete fractures (F-AMS). This paper introduces an Algebraic MultiScale method for simulation of flow in heterogeneous porous media with embedded discrete Fractures (F-AMS). First, multiscale coarse grids are independently constructed for both porous matrix and fracture networks. Then, a map between coarse- and fine-scale is obtained by algebraically computing basis functions with local support. In order to extend the localization assumption to the fractured media, four types of basis functions are investigated: (1) Decoupled-AMS, in which the two media are completely decoupled, (2) Frac-AMS and (3) Rock-AMS, which take into account only one-way transmissibilities, and (4) Coupled-AMS, in which the matrix and fracture interpolators are fully coupled. In order to ensure scalability, the F-AMS framework permits full flexibility in terms of the resolution of the fracture coarse grids. Numerical results are presented for two- and three-dimensional heterogeneous test cases. During these experiments, the performance of F-AMS, paired with ILU(0) as second-stage smoother in a convergent iterative procedure, is studied by monitoring CPU times and convergence rates. Finally, in order to investigate the scalability of the method, an extensive benchmark study is conducted, where a commercial algebraic multigrid solver is used as reference. The results show that, given an appropriate coarsening strategy, F-AMS is insensitive to severe fracture and matrix conductivity contrasts, as well as the length of the fracture networks. Its unique feature is that a fine-scale mass conservative flux field can be reconstructed after any iteration, providing efficient approximate solutions in time-dependent simulations.

The multiscale restriction smoothed basis method for fractured porous media (F-MsRSB). A novel multiscale method for multiphase flow in heterogeneous fractured porous media is devised. The discrete fine-scale system is described using an embedded fracture modeling approach, in which the heterogeneous rock (matrix) and highly-conductive fractures are represented on independent grids. Given this fine-scale discrete system, the method first partitions the fine-scale volumetric grid representing the matrix and the lower-dimensional grids representing fractures into independent coarse grids. Then, basis functions for matrix and fractures are constructed by restricted smoothing, which gives a flexible and robust treatment of complex geometrical features and heterogeneous coefficients. From the basis functions one constructs a prolongation operator that maps between the coarse- and fine-scale systems. The resulting method allows for general coupling of matrix and fracture basis functions, giving efficient treatment of a large variety of fracture conductivities. In addition, basis functions can be adaptively updated using efficient global smoothing strategies to account for multiphase flow effects. The method is conservative and because it is described and implemented in algebraic form, it is straightforward to employ it to both rectilinear and unstructured grids. Through a series of challenging test cases for single and multiphase flow, in which synthetic and realistic fracture maps are combined with heterogeneous petrophysical matrix properties, we validate the method and conclude that it is an efficient and accurate approach for simulating flow in complex, large-scale, fractured media.

Local-global splitting for spatiotemporal-adaptive multiscale methods. We present a novel spatiotemporal-adaptive Multiscale Finite Volume (MsFV) method, which is based on the natural idea that the global coarse-scale problem has longer characteristic time than the local fine-scale problems. As a consequence, the global problem can be solved with larger time steps than the local problems. In contrast to the pressure-transport splitting usually employed in the standard MsFV approach, we propose to start directly with a local–global splitting that allows to locally retain the original degree of coupling. This is crucial for highly non-linear systems or in the presence of physical instabilities. To obtain an accurate and efficient algorithm, we devise new adaptive criteria for global update that are based on changes of coarse-scale quantities rather than on fine-scale quantities, as it is routinely done before in the adaptive MsFV method. By means of a complexity analysis we show that the adaptive approach gives a noticeable speed-up with respect to the standard MsFV algorithm. In particular, it is efficient in case of large upscaling factors, which is important for multiphysics problems. Based on the observation that local time stepping acts as a smoother, we devise a self-correcting algorithm which incorporates the information from previous times to improve the quality of the multiscale approximation. We present results of multiphase flow simulations both for Darcy-scale and multiphysics (hybrid) problems, in which a local pore-scale description is combined with a global Darcy-like description. The novel spatiotemporal-adaptive multiscale method based on the local–global splitting is not limited to porous media flow problems, but it can be extended to any system described by a set of conservation equations.

Constrained pressure residual multiscale (CPR-MS) method for fully implicit simulation of multiphase flow in porous media We develop the first multiscale method for fully implicit (FIM) simulations of multiphase flow in porous media, namely CPR-MS method. Built on the FIM Jacobian matrix, the pressure system is obtained by employing a Constrained Pressure Residual (CPR) operator. Multiscale Finite Element (MSFE) and Finite Volume (MSFV) methods are then formulated algebraically to obtain efficient and accurate solutions of this pressure equation. The multiscale prediction stage (first-stage) is coupled with a corrector stage (second-stage) employed on the full system residual. The converged solution is enhanced through outer GMRES iterations preconditioned by these first and second stage operators. While the second-stage FIM stage is solved using a classical iterative solver, the multiscale stage is investigated in full detail. Several choices for fine-scale pre- and post-smoothing along with different choices of coarse-scale solvers are considered for a range of heterogeneous three-dimensional cases with capillarity and three-phase systems. The CPR-MS method is the first of its kind, and extends the applicability of the so-far developed multiscale methods (both MSFE and MSFV) to displacements with strong coupling terms.

A hierarchical fracture model for the iterative multiscale finite volume method An iterative multiscale finite volume (i-MSFV) method is devised for the simulation of multiphase flow in fractured porous media in the context of a hierarchical fracture modeling framework. Motivated by the small pressure change inside highly conductive fractures, the fully coupled system is split into smaller systems, which are then sequentially solved. This splitting technique results in only one additional degree of freedom for each connected fracture network appearing in the matrix system. It can be interpreted as an agglomeration of highly connected cells; similar as in algebraic multigrid methods. For the solution of the resulting algebraic system, an i-MSFV method is introduced. In addition to the local basis and correction functions, which were previously developed in this framework, local fracture functions are introduced to accurately capture the fractures at the coarse scale. In this multiscale approach there exists one fracture function per network and local domain, and in the coarse scale problem there appears only one additional degree of freedom per connected fracture network. Numerical results are presented for validation and verification of this new iterative multiscale approach for fractured porous media, and to investigate its computational efficiency. Finally, it is demonstrated that the new method is an effective multiscale approach for simulations of realistic multiphase flows in fractured heterogeneous porous media.

Adaptive fully implicit multi-scale finite-volume method for multi-phase flow and transport in heterogeneous porous media We describe a sequential fully implicit (SFI) multi-scale finite volume (MSFV) algorithm for nonlinear multi-phase flow and transport in heterogeneous porous media. The method extends the recently developed multiscale approach, which is based on an IMPES (IMplicit Pressure, Explicit Saturation) scheme [P. Jenny, S.H. Lee, H.A. Tchelepi, Adaptive multi-scale finite volume method for multi-phase flow and transport, Multiscale, Model. Simul. 3 (2005) 50-64]. That previous method was tested extensively and with a series of difficult test cases, where it was clearly demonstrated that the multiscale results are in excellent agreement with reference fine-scale solutions and that the computational efficiency of the MSFV algorithm is much higher than that of standard reservoir simulators. However, the level of detail and range of property variability included in reservoir characterization models continues to grow. For such models, the explicit treatment of the transport problem (i.e. saturation equations) in the IMPES-based multiscale method imposes severe restrictions on the time step size, and that can become the major computational bottleneck. Here we show how this problem is resolved with our sequential fully implicit (SFI) MSFV algorithm. Simulations of large (million cells) and highly heterogeneous problems show that the results obtained with the implicit multi-scale method are in excellent agreement with reference fine-scale solutions. Moreover, we demonstrate the robustness of the coupling scheme for nonlinear flow and transport, and we show that the MSFV algorithm offers great gains in computational efficiency compared to standard reservoir simulation methods.

A neural probabilistic language model A goal of statistical language modeling is to learn the joint probability function of sequences of words in a language. This is intrinsically difficult because of the curse of dimensionality: a word sequence on which the model will be tested is likely to be different from all the word sequences seen during training. Traditional but very successful approaches based on n-grams obtain generalization by concatenating very short overlapping sequences seen in the training set. We propose to fight the curse of dimensionality by learning a distributed representation for words which allows each training sentence to inform the model about an exponential number of semantically neighboring sentences. The model learns simultaneously (1) a distributed representation for each word along with (2) the probability function for word sequences, expressed in terms of these representations. Generalization is obtained because a sequence of words that has never been seen before gets high probability if it is made of words that are similar (in the sense of having a nearby representation) to words forming an already seen sentence. Training such large models (with millions of parameters) within a reasonable time is itself a significant challenge. We report on experiments using neural networks for the probability function, showing on two text corpora that the proposed approach significantly improves on state-of-the-art n-gram models, and that the proposed approach allows to take advantage of longer contexts.

A logic-based calculus of events Formal Logic can be used to represent knowledge of many kinds for many purposes. It can be used to formalize programs, program specifications, databases, legislation, and natural language in general. For many such applications of logic a representation of time is necessary. Although there have been several attempts to formalize the notion of time in classical first-order logic, it is still widely believed that classical logic is not adequate for the representation of time and that some form of non-classical Temporal Logic is needed. In this paper, we shall outline a treatment of time, based on the notion of event, formalized in the Horn clause subset of classical logic augmented with negation as failure. The resulting formalization is executable as a logic program. We use the term ''event calculus'' to relate it to the well-known ''situation calculus'' (McCarthy and Hayes 1969). The main difference between the two is conceptual: the situation calculus deals with global states whereas the event calculus deals with local events and time periods. Like the event calculus, the situation calculus can be formalized by means of Horn clauses augmented with negation by failure (Kowalski 1979). The main intended applications investigated in this paper are the updating of data- bases and narrative understanding. In order to treat both cases uniformly we have taken the view that an update consists of the addition of new knowledge to a knowledge base. The effect of explicit deletion of information in conventional databases is obtained without deletion by adding new knowledge about the end of the period of time for which the information holds.

A linear time algorithm for finding tree-decompositions of small treewidth In this paper, we give for constant k a linear-time algorithm that, given a graph G = (V, E), determines whether the treewidth of G is at most k and, if so, finds a tree-decomposition of G with treewidth at most k. A consequence is that every minor-closed class of graphs that does not contain all planar graphs has a linear-time recognition algorithm. Another consequence is that a similar result holds when we look instead for path-decompositions with pathwidth at mast some constant k.

Abductive Logic Programming Abstract: This paper is a survey and critical overview of recent work on the extensionof Logic Programming to perform Abductive Reasoning (Abductive LogicProgramming). We outline the general framework of Abduction and itsapplications to Knowledge Assimilation and Default Reasoning; and we introducean argumentation-theoretic approach to the use of abduction as aninterpretation for Negation as Failure. We also analyse the links betweenAbduction and the extension of Logic Programming obtained by...

Representing actions in equational logic programming A sound and complete approach for encoding the action description language A developed by M. Gelfond and V. Lifschitz in an equational logic program is given. Our results allow the comparison of the resource-oriented equational logic based approach and various other methods designed for reasoning about actions, most of them based on variants of the situation calculus, which were also related to the action description language recently. A non-trivial extension of A is proposed which allows to handle uncer- tainty in form of non-deterministic action descriptions, i.e. where actions may have alternative randomized efiects. It is described how the equational logic programming approach forms a sound and complete encoding of this extended action description language AND as well.

Energy efficiency through burstiness OS resource management policies traditionally employ buffering to "smooth out" fluctuations in resource demand. By minimizing the length of idle periods and the level of contention during non-idle periods, such smoothing tends to maximize overall throughput and minimize the latency of individual requests. For certain important devices, how- ever (disks, network interfaces, or even computational ele- ments), smoothing eliminates opportunities to save energy using low-power modes. As devices with such modes pro- liferate, and as energy efficiency becomes an increasingly important design consideration, we argue that OS poli- cies should be redesigned to increase burstiness for energy- sensitive devices. We are currently experimenting with techniques to in- crease the disk access pattern burstiness of the Linux op- erating system. Our results indicate that the deliberate cre- ation of bursty activity can save up to 78.5% of the energy consumed by a Hitachi DK23DA disk (in comparison with current policies), while simultaneously decreasing the neg- ative impact of disk congestion and spin-up latency on ap- plication performance.

Exploiting Web Log Mining for Web Cache Enhancement Improving the performance of the Web is a crucial requirement, since its popularity resulted in a large increase in the user perceived latency. In this paper, we describe a Web caching scheme that capitalizes on prefetching. Prefetching refers to the mechanism of deducing forthcoming page accesses of a client, based on access log information. Web log mining methods are exploited to provide effective prediction of Web-user accesses. The proposed scheme achieves a coordination between the two techniques (i.e., caching and prefetching). The prefetched documents are accommodated in a dedicated part of the cache, to avoid the drawback of incorrect replacement of requested documents. The requirements of the Web are taken into account, compared to the existing schemes for buffer management in database and operating systems. Experimental results indicate the superiority of the proposed method compared to the previous ones, in terms of improvement in cache performance.

Exploring Sequence Alignment Algorithms On Fpga-Based Heterogeneous Architectures With the rapid development of DNA sequencer, the rate of data generation is rapidly outpacing the rate at which it can be computationally processed. Traditional sequence alignment based on PC cannot fulfill the increasing demand. Accelerating the algorithm using FPGA provides the better performance compared to the other platforms. This paper will explain and classify the current sequence alignment algorithms. In addition, we analyze the different types of sequence alignment algorithms and present the taxonomy of FPGA-based sequence alignment implementations. This work will conclude the current solutions and provide a reference to further accelerating sequence alignment on a FPGA-based heterogeneous architecture.