Fairness Evaluation on German Credit Dataset
-0.13Equal OpportunityPRSMP
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| PRSMPModel=Neural Network (NN Model), p (perturbation probability)=0.33, r (ratio)=5, I (iterations)=102026.06 | -0.13 | -0.14 | -0.1 | 0.82 | 0 | |
| PRSMPModel=Neural Network (NN Model), p (perturbation probability)=0.25, r (ratio)=5, I (iterations)=102026.06 | -0.12 | -0.1 | -0.05 | 0.87 | 0 | |
| BaselineModel=Neural Network (NN Model)2026.06 | -0.08 | -0.06 | -0.02 | 0.9 | 0 | |
| PMiSModel=Neural Network (NN Model), p (perturbation probability)=0.50, I (iterations)=52026.06 | -0.08 | -0.07 | -0.02 | 0.92 | 0 | |
| PMiSModel=Neural Network (NN Model), p (perturbation probability)=0.25, I (iterations)=102026.06 | -0.07 | -0.12 | -0.11 | 0.84 | 0 | |
| PRSMPModel=Neural Network (NN Model), p (perturbation probability)=0.25, r (ratio)=5, I (iterations)=52026.06 | -0.07 | -0.08 | -0.05 | 0.91 | 0 | |
| PMiSModel=Neural Network (NN Model), p (perturbation probability)=0.25, I (iterations)=52026.06 | -0.05 | -0.05 | -0.01 | 0.94 | 0 | |
| PRSMPModel Architecture=LR Model, p=0.33, r=5, I=102026.06 | 0.01 | -0.01 | 0.01 | 0.98 | 0 | |
| PRSMPModel Architecture=LR Model, p=0.25, r=10, I=102026.06 | 0.02 | -0.03 | 0.02 | 0.95 | 0 | |
| PMiSModel Architecture=LR Model, p=0.33, I=252026.06 | 0.04 | 0.01 | 0.04 | 1.02 | 0 | |
| PRSMPModel Architecture=LR Model, p=0.25, r=5, I=102026.06 | 0.06 | 0.02 | -0.05 | 1.04 | 0 | |
| PRSMPModel Architecture=LR Model, p=0.25, r=5, I=52026.06 | 0.06 | 0.03 | 0.05 | 1.04 | 0 | |
| DomSwapModel=Neural Network (NN Model)2026.06 | 0.07 | 0.1 | 0.15 | 1.14 | 2.21 | |
| PMiSModel Architecture=LR Model, p=0.25, I=252026.06 | 0.08 | 0.05 | 0.08 | 1.08 | 0 | |
| PMiSModel Architecture=LR Model, p=0.25, I=102026.06 | 0.11 | 0.08 | 0.1 | 1.11 | 0 | |
| MixSwapModel=Neural Network (NN Model)2026.06 | 0.12 | 0.07 | 0.07 | 1.09 | 0 | |
| BaselineModel Architecture=LR Model2026.06 | 0.13 | 0.09 | 0.12 | 1.13 | 0 |