Identifying Optimal Trajectories on Acrobot v1 (top-5 trajectories)
73.2Average Trajectory LengthEntropy-Based
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Entropy-BasedTop-k ranking=52025.12 | 73.2 | -72.2 | |
| Bellman ErrorTop-k ranking=52025.12 | 70.8 | -69.8 | |
| Classic (ΔQ only)Top-k ranking=52025.12 | 70 | -69 | |
| Naive NormalizationTop-k ranking=52025.12 | 70 | -69 | |
| V-NormalizationTop-k ranking=52025.12 | 70 | -69 | |
| V-GoalTop-k ranking=52025.12 | 68.8 | -67.8 |