Off-policy policy selection on Ring environment 100 batches
0.84Max Utility DifferenceNORBU
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| NORBUCategory=Baseline, Risk level q=0.25, Batch size N={8, 16, 24, 32, 40, 48, 56}2021.05 | 0.84 | 0.03 | 0 | -0.71 | |
| EvCCategory=Selection Method, Criterion=a_0.25, Risk level q=0.25, Batch size N={8, 16, 24, 32, 40, 48, 56}2021.05 | 0.82 | 0.01 | 0 | -0.82 | |
| UnOCategory=Selection Method, Criterion=a_0.25, Risk level q=0.25, Batch size N={8, 16, 24, 32, 40, 48, 56}2021.05 | 0.82 | -0.26 | -0.27 | -0.96 | |
| BCRCategory=Baseline, Risk level q=0.25, Batch size N={8, 16, 24, 32, 40, 48, 56}2021.05 | 0.74 | -0.01 | 0 | -0.82 | |
| UnOCategory=Selection Method, Criterion=phi_0.25, Risk level q=0.25, Batch size N={8, 16, 24, 32, 40, 48, 56}2021.05 | 0.72 | -0.27 | -0.33 | -0.96 | |
| EvCCategory=Selection Method, Criterion=phi_0.25, Risk level q=0.25, Batch size N={8, 16, 24, 32, 40, 48, 56}2021.05 | 0.71 | -0.04 | 0 | -0.82 | |
| SPIBBCategory=Baseline, Risk level q=0.25, Batch size N={8, 16, 24, 32, 40, 48, 56}2021.05 | 0.61 | -0.29 | -0.31 | -0.78 | |
| BOPAHCategory=Baseline, Risk level q=0.25, Batch size N={8, 16, 24, 32, 40, 48, 56}2021.05 | 0.48 | -0.28 | -0.34 | -0.68 |