Offline Multi-Agent Reinforcement Learning with Implicit Global-to-Local Value Regularization
It addresses the problem of learning effective policies from offline datasets in multi-agent systems, which is crucial for applications like robotics and gaming, but the approach appears incremental as it builds on existing offline MARL methods with a novel regularization technique.
The paper tackles the challenge of offline multi-agent reinforcement learning by introducing OMIGA, an algorithm that uses implicit global-to-local value regularization to bridge multi-agent value decomposition with offline policy learning, achieving superior performance over state-of-the-art methods in tasks like MuJoCo and StarCraft II.
Offline reinforcement learning (RL) has received considerable attention in recent years due to its attractive capability of learning policies from offline datasets without environmental interactions. Despite some success in the single-agent setting, offline multi-agent RL (MARL) remains to be a challenge. The large joint state-action space and the coupled multi-agent behaviors pose extra complexities for offline policy optimization. Most existing offline MARL studies simply apply offline data-related regularizations on individual agents, without fully considering the multi-agent system at the global level. In this work, we present OMIGA, a new offline m ulti-agent RL algorithm with implicit global-to-local v alue regularization. OMIGA provides a principled framework to convert global-level value regularization into equivalent implicit local value regularizations and simultaneously enables in-sample learning, thus elegantly bridging multi-agent value decomposition and policy learning with offline regularizations. Based on comprehensive experiments on the offline multi-agent MuJoCo and StarCraft II micro-management tasks, we show that OMIGA achieves superior performance over the state-of-the-art offline MARL methods in almost all tasks.