Multi-objective Reinforcement Learning on MuJoCo Hopper-3
3.889Hypervolume (HV)PA2D-MORL
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| PA2D-MORLObjectives=32026.03 | 3.889 | 2.1 | |
| PGMORLObjectives=32026.03 | 3.766 | 3.2 | |
| PA2D-ablatedAblation=without PA-FT during training, Objectives=32026.03 | 3.759 | 10.6 | |
| MOEA/DModification=modified version based on evolutionary framework, Objectives=32026.03 | 3.681 | 64.2 |