Regression on houseelectric (test)
0.029RMSEBTE
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| BTEn=1311539, d=11, GPU=Tesla V100-SXM2-16GB, Feature transformation=ECDF, Precision=Pnew2022.10 | 0.029 | -2.569 | — | — | |
| BTn=1311539, d=11, GPU=Tesla V100-SXM2-16GB, Feature transformation=ECDF, Precision=Pnew2022.10 | 0.029 | -2.492 | — | — | |
| SGPRInducing Points=4000, Time (hours)=50, GPUs=12022.06 | 0.043 | -1.717 | — | — | |
| SGPRInducing Points=3000, Time (hours)=38.68, GPUs=12022.06 | 0.044 | -1.696 | — | — | |
| SGPRInducing Points=2000, Time (hours)=18.03, GPUs=12022.06 | 0.046 | -1.651 | — | — | |
| SGPRInducing Points=1000, Time (hours)=5.01, GPUs=12022.06 | 0.048 | -1.602 | — | — | |
| Iterative GPlengthscale=shared across dimensions, Time (hours)=79.96, GPUs=82022.06 | 0.05 | — | — | — | |
| Iterative GPlengthscale=per dimension, Time (hours)=1.55, GPUs=82022.06 | 0.054 | -0.207 | — | — | |
| MATÉRN (BBMM)n=1311539, d=11, GPU=Tesla V100-SXM2-32GB, Feature transformation=ECDF, Precision=Pnew2022.10 | 0.055 | -0.152 | — | — | |
| SGPR2023.05 | 0.067 | — | — | — | |
| Simplex-GPInterpolation=Simplicial2023.05 | 0.078 | — | — | 0.186 | |
| Simplex-GPn=1311539, d=11, GPU=Tesla V100-SXM2-32GB, Feature transformation=ECDF, Precision=Pnew2022.10 | 0.079 | 0.756 | — | — | |
| SVGPn=1311539, d=11, GPU=Tesla V100-SXM2-32GB, Feature transformation=ECDF, Precision=Pnew2022.10 | 0.084 | -1.01 | — | — | |
| Sparse-gridGrid Type=Sparse2023.05 | 0.088 | — | — | 0.047 | |
| Dense-gridGrid Type=Dense2023.05 | 0.17 | — | — | 0.263 |