Regression on Energy UCI (test)
0.04RMSELA-BNN
Evaluation results
| Method | Links | ||
|---|---|---|---|
| LA-BNN2026.04 | 0.04 | 0.47 | |
| LA-TNKM2026.04 | 0.05 | 0.2 | |
| SP-BTN2026.04 | 0.43 | 0.07 | |
| GPR + HMCM (Number of inducing points)=N/A (Full GP), q(theta) (Posterior approximation configuration)=-2022.11 | 0.441 | — | |
| CIBERpartition=SECOND2023.06 | 0.447 | — | |
| CIBERpartition=LAST2023.06 | 0.447 | — | |
| SGPR + HMCM (Number of inducing points)=100, q(theta) (Posterior approximation configuration)=free-form2022.11 | 0.469 | — | |
| VSD2021.02 | 0.47 | — | |
| VSD2023.06 | 0.47 | — | |
| SGPRM (Number of inducing points)=100, q(theta) (Posterior approximation configuration)=delta2022.11 | 0.477 | — | |
| FBGPM (Number of inducing points)=100, q(theta) (Posterior approximation configuration)=free-form (Z, theta)2022.11 | 0.49 | — | |
| SVGPM (Number of inducing points)=100, q(theta) (Posterior approximation configuration)=delta2022.11 | 0.5 | — | |
| VMG2021.02 | 0.54 | — | |
| SLANG2021.02 | 0.64 | — | |
| VD2021.02 | 0.64 | — | |
| Sparse-gridGrid type=Sparse2023.05 | 0.715 | — | |
| JointHMCM (Number of inducing points)=100, q(theta) (Posterior approximation configuration)=free-form2022.11 | 0.755 | — | |
| BBB2021.02 | 0.97 | — | |
| MF-BTN2026.04 | 0.98 | 0.1 | |
| MCD2021.02 | 1.08 | — | |
| MCD2023.06 | 1.08 | — | |
| Dense-gridGrid type=Dense2023.05 | 1.333 | — | |
| SGPR2023.05 | 1.509 | — | |
| PCA+VImode=SI2023.06 | 1.587 | — | |
| SWAG2023.06 | 1.594 | — | |
| PCA+ESSmode=SI2023.06 | 1.598 | — | |
| SGD2023.06 | 1.602 | — | |
| NE-GMM2026.03 | 1.806 | — | |
| SampleNet2026.03 | 1.866 | — | |
| β-NLL2026.03 | 1.944 | — | |
| NGBoost2026.03 | 1.952 | — | |
| MDN2026.03 | 1.985 | — | |
| D.E2021.02 | 2.09 | — | |
| Ensemble-NN2026.03 | 2.187 | — | |
| MNF2021.02 | 2.38 | — | |
| Simplex-GPInterpolation=Simplicial2023.05 | 3.076 | — | |
| SKIP2023.05 | 5.762 | — |