Regression on UCI-DIR Abalone (test)
4.158MAE (All)SRL
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| SRLBackbone=3layer MLP, Feature Dim=128, Learning Rate=1e-3, Batch Size=256, Temperature (τ)=0.1, Momentum (α)=0.9, N=1000, λe=1e-2, λh=1e-22025.03 | 4.158 | 0.892 | 2.423 | 7.191 | |
| SRLArchitecture=MLP (d-20-30-10-1)2025.03 | 4.16 | 0.89 | 2.42 | 7.19 | |
| RankSimArchitecture=MLP (d-20-30-10-1)2025.03 | 4.33 | 0.98 | 2.59 | 7.42 | |
| RankSimBackbone=3layer MLP, Feature Dim=128, Learning Rate=1e-3, Batch Size=2562025.03 | 4.332 | 0.975 | 2.591 | 7.421 | |
| VANILLABackbone=3layer MLP, Feature Dim=128, Learning Rate=1e-3, Batch Size=2562025.03 | 4.567 | 0.878 | 2.646 | 7.967 | |
| VANILLAArchitecture=MLP (d-20-30-10-1)2025.03 | 4.57 | 0.88 | 2.65 | 7.97 | |
| LDS + FDSBackbone=3layer MLP, Feature Dim=128, Learning Rate=1e-3, Batch Size=2562025.03 | 5.087 | 0.904 | 3.261 | 9.261 | |
| LDS + FDSArchitecture=MLP (d-20-30-10-1)2025.03 | 5.09 | 0.9 | 3.26 | 9.26 | |
| BalancedMSEBackbone=3layer MLP, Feature Dim=128, Learning Rate=1e-3, Batch Size=2562025.03 | 5.366 | 2.135 | 2.659 | 9.368 | |
| BalancedMSEArchitecture=MLP (d-20-30-10-1)2025.03 | 5.37 | 2.14 | 2.66 | 9.37 | |
| Ordinal EntropyArchitecture=MLP (d-20-30-10-1)2025.03 | 6.77 | 2.31 | 4.01 | 11.61 | |
| Ordinal EntropyBackbone=3layer MLP, Feature Dim=128, Learning Rate=1e-3, Batch Size=2562025.03 | 6.774 | 2.314 | 4.013 | 11.61 |