Off-policy prediction on Boyan chain
0.166Tail-average RMSERETD
Evaluation Results
| Method | Links | |
|---|---|---|
| RETDalpha=0.01, tail-average window=final 10% of update steps2026.05 | 0.166 | |
| TDRCProtocol=at the last step, Runs=100 independent2026.05 | 0.1667 | |
| HTDProtocol=at the last step, Runs=100 independent2026.05 | 0.1669 | |
| ETDProtocol=at the last step, Runs=100 independent2026.05 | 0.1669 | |
| TDProtocol=at the last step, Runs=100 independent2026.05 | 0.1677 | |
| STHTD-MPProtocol=at the last step, Runs=100 independent2026.05 | 0.1687 | |
| ETDalpha=0.01, tail-average window=final 10% of update steps2026.05 | 0.172 | |
| GTD2Protocol=at the last step, Runs=100 independent2026.05 | 0.1726 | |
| STHTDProtocol=at the last step, Runs=100 independent2026.05 | 0.177 | |
| GTD2-MPProtocol=at the last step, Runs=100 independent2026.05 | 0.2747 | |
| TDCProtocol=at the last step, Runs=100 independent2026.05 | 0.2777 | |
| TETDalpha=0.01, tail-average window=final 10% of update steps2026.05 | 0.282 | |
| TDalpha=0.01, tail-average window=final 10% of update steps2026.05 | 4.666 | |
| TDRCalpha=0.01, tail-average window=final 10% of update steps2026.05 | 4.832 | |
| TDCalpha=0.01, tail-average window=final 10% of update steps2026.05 | 12.17 | |
| GTD2alpha=0.01, tail-average window=final 10% of update steps2026.05 | 12.4 |