Synthetic Shortest Path on Synthetic Shortest Path SP2 (test)
10.07Regret (%)SPO
Evaluation results
| Method | Links | |
|---|---|---|
| SPODFL Technique=Smart Predict-then-Optimize, Regularization=None2026.01 | 10.07 | |
| MSELoss Function=Mean-Squared Error2026.01 | 10.12 | |
| SPO-rnDFL Technique=Smart Predict-then-Optimize, Regularization=r^n2026.01 | 10.12 | |
| SPO-rpDFL Technique=Smart Predict-then-Optimize, Regularization=r^p2026.01 | 10.4 | |
| DPO-rnDFL Technique=Differentiable Perturbed Optimizers, Regularization=r^n2026.01 | 10.55 | |
| DBB-rpDFL Technique=Differentiable Black-Box, Regularization=r^p2026.01 | 10.83 | |
| DPO-rpDFL Technique=Differentiable Perturbed Optimizers, Regularization=r^p2026.01 | 11.18 | |
| DBB-rnDFL Technique=Differentiable Black-Box, Regularization=r^n2026.01 | 11.32 | |
| DPODFL Technique=Differentiable Perturbed Optimizers, Regularization=None2026.01 | 11.83 | |
| DBBDFL Technique=Differentiable Black-Box, Regularization=None2026.01 | 11.92 |