Learning Cooperative Multi-Agent Policies with Partial Reward Decoupling
This addresses a key scaling obstacle for multi-agent reinforcement learning, offering an incremental improvement over existing methods like COMA.
The paper tackles the credit assignment problem in multi-agent reinforcement learning by introducing partial reward decoupling (PRD), which decomposes cooperative tasks into subproblems to simplify credit assignment, resulting in improved data efficiency, stability, and asymptotic performance across various tasks.
One of the preeminent obstacles to scaling multi-agent reinforcement learning to large numbers of agents is assigning credit to individual agents' actions. In this paper, we address this credit assignment problem with an approach that we call \textit{partial reward decoupling} (PRD), which attempts to decompose large cooperative multi-agent RL problems into decoupled subproblems involving subsets of agents, thereby simplifying credit assignment. We empirically demonstrate that decomposing the RL problem using PRD in an actor-critic algorithm results in lower variance policy gradient estimates, which improves data efficiency, learning stability, and asymptotic performance across a wide array of multi-agent RL tasks, compared to various other actor-critic approaches. Additionally, we relate our approach to counterfactual multi-agent policy gradient (COMA), a state-of-the-art MARL algorithm, and empirically show that our approach outperforms COMA by making better use of information in agents' reward streams, and by enabling recent advances in advantage estimation to be used.