Bayesian role discovery for multi-agent reinforcement learning
Aaron Wilson, Alan Fern, Prasad V. Tadepalli · Adaptive Agents and Multi-Agents Systems · 2010
In this paper we develop a Bayesian policy search approach for Multi-Agent RL (MARL), which is model-free and allows for priors on policy parameters. We present a novel optimization algorithm based on hybrid MCMC, which leverages both the prior and gradient information estimated from trajectories. Our experiments demonstrate the automatic discovery of roles through reinforcement learning in a real-time strategy game.