Multi-objective Multi-agent Reinforcement Learning on SMAC 2s3z
7,371.7852HypervolumeMO-MIX
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| MO-MIXScenario=2s3z, Training steps=5 million, Sampling interval of preference=0.01252026.02 | 7,371.7852 | 18.4 | 7.8697 | 203.704 | |
| Outer-loop QMIXScenario=2s3z, Training steps=41 million, Rounds=412026.02 | 6,226.9744 | 10 | 33.5992 | 1,509.6655 |