Tabular Regression on Energy UCI 10% outlier contamination
1.89RMSEEM
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| EMOutlier Contamination=10%, Feature Mapping=Random Fourier Features (RFF), Dimensions (d)=256, Trials=10, Model=Heteroscedastic SBL2026.05 | 1.89 | 29.4 | 91.2 | |
| ℓ2-IRLSOutlier Contamination=10%, Feature Mapping=Random Fourier Features (RFF), Dimensions (d)=256, Trials=10, Model=Heteroscedastic SBL2026.05 | 1.89 | 39.5 | 91 | |
| MacKayOutlier Contamination=10%, Feature Mapping=Random Fourier Features (RFF), Dimensions (d)=256, Trials=10, Model=Heteroscedastic SBL2026.05 | 1.97 | 6.7 | 88.4 | |
| HuberOutlier Contamination=10%, Feature Mapping=Random Fourier Features (RFF), Dimensions (d)=256, Trials=102026.05 | 2.14 | 100 | 83.5 | |
| Student-tOutlier Contamination=10%, Feature Mapping=Random Fourier Features (RFF), Dimensions (d)=256, Trials=102026.05 | 2.23 | 100 | 92.5 | |
| Grad. (Primal)Outlier Contamination=10%, Feature Mapping=Random Fourier Features (RFF), Dimensions (d)=256, Trials=10, Model=Heteroscedastic SBL2026.05 | 2.28 | 11.8 | 64.4 | |
| Grad. (Dual)Outlier Contamination=10%, Feature Mapping=Random Fourier Features (RFF), Dimensions (d)=256, Trials=10, Model=Heteroscedastic SBL2026.05 | 2.28 | 11.1 | 64.4 | |
| ℓ1-IRLSOutlier Contamination=10%, Feature Mapping=Random Fourier Features (RFF), Dimensions (d)=256, Trials=10, Model=Heteroscedastic SBL2026.05 | 2.48 | 18.5 | 95.4 | |
| RidgeOutlier Contamination=10%, Feature Mapping=Random Fourier Features (RFF), Dimensions (d)=256, Trials=102026.05 | 2.79 | 96 | 100 | |
| GPOutlier Contamination=10%, Feature Mapping=Random Fourier Features (RFF), Dimensions (d)=256, Trials=102026.05 | 2.85 | 9.2 | 100 | |
| OLSOutlier Contamination=10%, Feature Mapping=Random Fourier Features (RFF), Dimensions (d)=256, Trials=102026.05 | 5.73 | 100 | 100 |