Cooperative Multi-Agent Reinforcement Learning on SustainGym BuildingEnv season_2 (test)
91.6Normalized Episodic Returnrobust QMIX
Evaluation results
| Method | Links | |
|---|---|---|
| robust QMIXFactorization Method=QMIX, Uncertainty Set=TV-uncertainty2026.02 | 91.6 | |
| robust QMIXFactorization Method=QMIX, Uncertainty Set=rho-contamination2026.02 | 91.1 | |
| robust VDNFactorization Method=VDN, Uncertainty Set=TV-uncertainty2026.02 | 89.8 | |
| QMIXFactorization Method=QMIX, Robustness=Non-robust2026.02 | 89.5 | |
| VDNFactorization Method=VDN, Robustness=Non-robust2026.02 | 87.7 | |
| robust VDNFactorization Method=VDN, Uncertainty Set=rho-contamination2026.02 | 86.9 | |
| robust QTRANFactorization Method=QTRAN, Uncertainty Set=TV-uncertainty2026.02 | 86.1 | |
| robust QTRANFactorization Method=QTRAN, Uncertainty Set=rho-contamination2026.02 | 82.5 | |
| QTRANFactorization Method=QTRAN, Robustness=Non-robust2026.02 | 81.6 | |
| GroupDRFactorization Method=VDN, Robustness=baseline2026.02 | 62.4 | |
| GroupDRFactorization Method=QTRAN, Robustness=baseline2026.02 | 50.8 | |
| GroupDRFactorization Method=QMIX, Robustness=baseline2026.02 | 49.9 |