A Regularized Opponent Model with Maximum Entropy Objective
This work addresses performance improvement in cooperative multi-agent reinforcement learning, presenting a novel theoretical framework with practical algorithms.
The paper tackles multi-agent reinforcement learning by reformulating it as probabilistic inference, introducing ROMMEO, a method that improves agent performance in cooperative games, with empirical results showing it outperforms strong baselines in iterated matrix and differential games.
In a single-agent setting, reinforcement learning (RL) tasks can be cast into an inference problem by introducing a binary random variable o, which stands for the "optimality". In this paper, we redefine the binary random variable o in multi-agent setting and formalize multi-agent reinforcement learning (MARL) as probabilistic inference. We derive a variational lower bound of the likelihood of achieving the optimality and name it as Regularized Opponent Model with Maximum Entropy Objective (ROMMEO). From ROMMEO, we present a novel perspective on opponent modeling and show how it can improve the performance of training agents theoretically and empirically in cooperative games. To optimize ROMMEO, we first introduce a tabular Q-iteration method ROMMEO-Q with proof of convergence. We extend the exact algorithm to complex environments by proposing an approximate version, ROMMEO-AC. We evaluate these two algorithms on the challenging iterated matrix game and differential game respectively and show that they can outperform strong MARL baselines.