Counterfactual Explanation Generation on German Credit Protocol A B=4
88.23SparsityDICE-R
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| DICE-RClassifier=Logistic Regression, Protocol=A, B=42026.02 | 88.23 | 15.29 | 26.06 | 100 | 100 | 90.81 | |
| COUNTERFLOWNETClassifier=Logistic Regression, Protocol=A, B=42026.02 | 69.09 | 41.26 | 51.65 | 99.48 | 100 | 100 | |
| L2CClassifier=Logistic Regression, Protocol=A, B=42026.02 | 61.35 | 37.31 | 46.39 | 100 | 100 | 99.06 | |
| COPAClassifier=Logistic Regression, Protocol=A, B=42026.02 | 57.88 | 18.88 | 28.47 | 44 | 44 | 84.31 | |
| DICE-GClassifier=Logistic Regression, Protocol=A, B=42026.02 | 43.45 | 37.56 | 40.29 | 62.87 | 90.24 | 56.66 | |
| MCCEClassifier=Logistic Regression, Protocol=A, B=42026.02 | 28.76 | 33.4 | 30.91 | 48.74 | 100 | 58.76 |