Federated Softmax Policy Gradient on Federated Reinforcement Learning with Heterogeneous Dynamics
1Communication ComplexityFEDSVRPG-M
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| FEDSVRPG-MObjective=Unregularized, Convergence Target=ϵ-stationary point2025.05 | 1 | — | — | 1 | |
| FEDHAPG-MObjective=Unregularized, Convergence Target=ϵ-stationary point2025.05 | 1 | — | — | 1 | |
| S-FedPGObjective=Unregularized, Convergence Target=ϵ-optimal solution2025.05 | 1 | — | — | 1 | |
| RS-FedPGObjective=Entropy-regularized, Convergence Target=ϵ-optimal solution2025.05 | 1 | — | — | 1 |