Algorithmic Recourse on Attrition (10-fold cross-validation)
0.02Recourse Cost (MPS)AReS
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| AReSClassifier Model=XGBoost2026.05 | 0.02 | 0.97 | 0.99 | 11.68 | |
| SOGARClassifier Model=DNN2026.05 | 0.05 | 0.01 | 0.05 | 250 | |
| AReSClassifier Model=DNN2026.05 | 0.23 | 0.23 | 0.46 | 139.3 | |
| SOGARClassifier Model=XGBoost2026.05 | 0.23 | 0.02 | 0.26 | 254.1 | |
| CETClassifier Model=LightGBM2026.05 | 0.26 | 0.58 | 0.84 | 286.7 | |
| CETClassifier Model=DNN2026.05 | 0.3 | 0.1 | 0.4 | 198.7 | |
| SOGARClassifier Model=LightGBM2026.05 | 0.3 | 0.07 | 0.37 | 356.4 | |
| T-CREx0.9Classifier Model=XGBoost2026.05 | 0.33 | 0.72 | 1.1 | 7.99 | |
| T-CREx0.9Classifier Model=DNN2026.05 | 0.35 | 0.77 | 1.13 | 7 | |
| T-CREx0.9Classifier Model=LightGBM2026.05 | 0.37 | 0.8 | 1.16 | 7.08 | |
| CETClassifier Model=XGBoost2026.05 | 0.4 | 0.17 | 0.57 | 342.6 | |
| AReSClassifier Model=LightGBM2026.05 | 0.44 | 0.44 | 0.88 | 307.4 | |
| T-CREx0.99Classifier Model=XGBoost2026.05 | 0.48 | 0.64 | 1.12 | 14.13 | |
| GLANCEClassifier Model=DNN2026.05 | 0.54 | 0.01 | 0.54 | 11.66 | |
| T-CREx0.99Classifier Model=DNN2026.05 | 0.56 | 0.65 | 1.21 | 176.32 | |
| GLANCEClassifier Model=XGBoost2026.05 | 0.56 | 0.02 | 0.58 | 11.7 | |
| T-CREx0.99Classifier Model=LightGBM2026.05 | 0.58 | 0.84 | 1.42 | 112.67 | |
| GLANCEClassifier Model=LightGBM2026.05 | 0.62 | 0.08 | 0.7 | 12.17 | |
| GLOBE-CEClassifier Model=XGBoost2026.05 | 0.75 | 0 | 0.75 | 5.9 | |
| GLOBE-CEClassifier Model=LightGBM2026.05 | 0.79 | 0.01 | 0.8 | 7.96 | |
| GLOBE-CEClassifier Model=DNN2026.05 | 0.81 | 0.01 | 0.81 | 7.6 |