Classification on Breast Cancer (UCI) (test)
0.117NLLDIMS
Evaluation results
| Method | Links | |
|---|---|---|
| DIMSNumber of samples (S)=30, Number of seeds=5, backbone=3-layer network, hidden units=16, activations=SiLU, dropout=50%2026.05 | 0.117 | |
| RIEM-LAPrior precision=optimized, Number of MC samples=302023.06 | 0.176 | |
| RLANumber of samples (S)=30, Number of seeds=5, backbone=3-layer network, hidden units=16, activations=SiLU, dropout=50%2026.05 | 0.185 | |
| LINEARIZED RIEM-LAPrior precision=optimized, Number of MC samples=302023.06 | 0.202 | |
| MAPPrior precision=optimized, Number of MC samples=302023.06 | 0.225 | |
| LINEARIZED LAPrior precision=optimized, Number of MC samples=302023.06 | 0.327 | |
| VANILLA LAPrior precision=optimized, Number of MC samples=302023.06 | 0.594 | |
| LINLANumber of samples (S)=30, Number of seeds=5, backbone=3-layer network, hidden units=16, activations=SiLU, dropout=50%2026.05 | 1.662 | |
| LANumber of samples (S)=30, Number of seeds=5, backbone=3-layer network, hidden units=16, activations=SiLU, dropout=50%2026.05 | 10.084 |