Current performance
Accuracy0.697-0.135 vs reference
Macro F10.688-0.134 vs reference
Log loss0.876+0.269 vs reference
ECE0.169+0.075 vs reference
Distribution shift
Prediction JS divergence0.0007
Label JS divergence0.0001
Confidence PSI0.0412
Worst-first slice analysis
| Slice | Value | Count | Accuracy | Gap | ECE |
|---|---|---|---|---|---|
| region | west | 208 | 0.625 | -0.072 | 0.225 |
| device | mobile | 304 | 0.671 | -0.026 | 0.173 |
| device | desktop | 296 | 0.723 | +0.026 | 0.165 |
| region | north | 189 | 0.730 | +0.033 | 0.163 |
| region | south | 203 | 0.739 | +0.042 | 0.118 |
Class-level performance
| Class | Support | Precision | Recall | F1 |
|---|---|---|---|---|
| approve | 269 | 0.777 | 0.714 | 0.744 |
| reject | 149 | 0.608 | 0.698 | 0.650 |
| review | 182 | 0.670 | 0.670 | 0.670 |
Selective prediction
10%0.0% risk
20%14.2% risk
30%23.3% risk
40%32.1% risk
50%36.0% risk
60%35.6% risk
70%34.8% risk
80%34.8% risk
90%33.7% risk
100%30.3% risk