Regression on POL
2.555Log LikelihoodCIBER
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| CIBERcollapsed parameter set=second-to-last layer2023.06 | 2.555 | — | — | — | — | |
| CIBERcollapsed parameter set=last layer2023.06 | 2.506 | — | — | — | — | |
| PCA+VIvariant=SI2023.06 | 1.764 | — | — | — | — | |
| SWAG2023.06 | 1.533 | — | — | — | — | |
| SGD2023.06 | 1.073 | — | — | — | — | |
| OrthVGP2023.06 | 0.159 | — | — | — | — | |
| PCA+ESSvariant=SI2023.06 | -0.185 | — | — | — | — | |
| NL2023.06 | -2.84 | — | — | — | — | |
| CARTNumber of features (Nf)=26, Number of training samples (Ns)=105002026.05 | — | 11.02 | — | — | — | |
| CGNoise Level=Standard2023.06 | — | 0.08 | — | 2.18 | -1.17 | |
| CGNoise Level=Low2023.06 | — | — | 0.16 | — | — | |
| DGT-LinearHeight=5, Number of features (Nf)=26, Number of training samples (Ns)=105002026.05 | — | 8.03 | — | — | — | |
| DTSemNet STEHeight=5, Number of features (Nf)=26, Number of training samples (Ns)=105002026.05 | — | 7.62 | — | — | — | |
| DTSemNet Top-kHeight=5, Number of features (Nf)=26, Number of training samples (Ns)=105002026.05 | — | 6.61 | — | — | — | |
| SGD-GPNoise Level=Standard2023.06 | — | 0.13 | — | 3.51 | -0.7 | |
| SGD-GPNoise Level=Low2023.06 | — | — | 0.13 | — | — | |
| SVGPNoise Level=Standard2023.06 | — | 0.1 | — | 21.2 | -0.71 |