explain
Browse files
app.py
CHANGED
@@ -225,13 +225,12 @@ With no need for jargon, SSDS delivers tangible value to our fintech operations.
|
|
225 |
|
226 |
gr.Markdown("""
|
227 |
|
228 |
-
**Context:**
|
229 |
-
This analysis is derived from an XGBoost regression model designed to predict house prices. The model utilizes features such as **dist_subway, age, lat, long,** and **dist_stores**.
|
230 |
|
231 |
-
Full dataset at the bottom of this tab
|
232 |
|
|
|
|
|
233 |
Explain by Context
|
234 |
-
|
235 |
- Sometimes, understanding why an individual defaults requires shifting to a credit-healthy background, altering the baseline E[f(x) | credit healthy] using interventional feature perturbation ([source](https://arxiv.org/pdf/2006.16234.pdf)).
|
236 |
|
237 |
[UCI Machine Learning Repository - Credit Default Dataset](https://www.kaggle.com/datasets/uciml/default-of-credit-card-clients-dataset)
|
@@ -249,10 +248,15 @@ f(x) in probability for logistic regression objective using XGBoost
|
|
249 |
- LIMIT_BAL signifies the amount of given credit in NT dollars (includes individual and family/supplementary credit).
|
250 |
- BILL_AMT1 indicates the bill statement amount in September, 2005 (NT dollar).
|
251 |
|
252 |
-
|
253 |
-
|
|
|
|
|
|
|
|
|
|
|
254 |
Explain by Dataset
|
255 |
-
|
256 |
- Below are explanation in typical background E[f(x)]
|
257 |
|
258 |
![Summary](file=./xgb/data.png)
|
@@ -265,7 +269,7 @@ Explain by Dataset
|
|
265 |
|
266 |
|
267 |
Explain by Feature
|
268 |
-
|
269 |
![Partial Dependence](file=./xgb/feature.png)
|
270 |
|
271 |
**Observations:**
|
@@ -274,7 +278,7 @@ Explain by Feature
|
|
274 |
|
275 |
|
276 |
Explain by Record
|
277 |
-
|
278 |
![Force](file=./xgb/record.png)
|
279 |
|
280 |
**Contribution to Price:**
|
@@ -282,7 +286,7 @@ Explain by Record
|
|
282 |
- **Age** follows as the second significant contributor.
|
283 |
|
284 |
Explain by Instance
|
285 |
-
|
286 |
![Dependence](file=./xgb/instance.png)
|
287 |
|
288 |
**Insights:**
|
|
|
225 |
|
226 |
gr.Markdown("""
|
227 |
|
|
|
|
|
228 |
|
|
|
229 |
|
230 |
+
CREDIT DEFAULT RISK INTERPRETATION
|
231 |
+
=======================
|
232 |
Explain by Context
|
233 |
+
----------
|
234 |
- Sometimes, understanding why an individual defaults requires shifting to a credit-healthy background, altering the baseline E[f(x) | credit healthy] using interventional feature perturbation ([source](https://arxiv.org/pdf/2006.16234.pdf)).
|
235 |
|
236 |
[UCI Machine Learning Repository - Credit Default Dataset](https://www.kaggle.com/datasets/uciml/default-of-credit-card-clients-dataset)
|
|
|
248 |
- LIMIT_BAL signifies the amount of given credit in NT dollars (includes individual and family/supplementary credit).
|
249 |
- BILL_AMT1 indicates the bill statement amount in September, 2005 (NT dollar).
|
250 |
|
251 |
+
|
252 |
+
HOME PRICE INTERPRETATION
|
253 |
+
=======================
|
254 |
+
This analysis is derived from an XGBoost regression model designed to predict house prices. The model utilizes features such as **dist_subway, age, lat, long,** and **dist_stores**.
|
255 |
+
|
256 |
+
Full dataset at the bottom of this tab
|
257 |
+
|
258 |
Explain by Dataset
|
259 |
+
----------
|
260 |
- Below are explanation in typical background E[f(x)]
|
261 |
|
262 |
![Summary](file=./xgb/data.png)
|
|
|
269 |
|
270 |
|
271 |
Explain by Feature
|
272 |
+
----------
|
273 |
![Partial Dependence](file=./xgb/feature.png)
|
274 |
|
275 |
**Observations:**
|
|
|
278 |
|
279 |
|
280 |
Explain by Record
|
281 |
+
----------
|
282 |
![Force](file=./xgb/record.png)
|
283 |
|
284 |
**Contribution to Price:**
|
|
|
286 |
- **Age** follows as the second significant contributor.
|
287 |
|
288 |
Explain by Instance
|
289 |
+
----------
|
290 |
![Dependence](file=./xgb/instance.png)
|
291 |
|
292 |
**Insights:**
|