Multi-class Classification on LPMC holdout (test)
0.674Cross-Entropy LossAPLR
Evaluation results
| Method | Links | |
|---|---|---|
| APLRType=Piecewise-linear GAM2026.04 | 0.674 | |
| ParamBoostImplementation=JAX2026.04 | 0.674 | |
| NAMArchitecture=MLP (64, 64, 32), Optimizer=AdamW2026.04 | 0.676 | |
| EBMType=Tree-based GAM2026.04 | 0.677 | |
| NBMArchitecture=100 shared basis functions, Optimizer=AdamW2026.04 | 0.677 | |
| Linear/Logistic regressionImplementation=scikit-learn2026.04 | 0.731 | |
| MGCVImplementation=PyMGCV, Bases=CubicSpline2026.04 | 1.344 |