Policy Learning on Tabular Discounted MDPs Generative Setting
2Upper Bound (Complexity)MB-RS-QVI
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| MB-RS-QVIObjective=ERM, Risk parameter (beta)=beta != 02025.05 | 2 | 2 | |
| Risk-neutralObjective=Risk-neutral, Risk parameter (beta)=02025.05 | 2 | 2 |