Decentralized POMDPs provide a rigorous framework for multi-agent decision-theoretic planning. However, their high complexity has limited scalability. In this work, we present a promising new class of algorithms based on probabilistic inference for infinite-horizon ND-POMDPs - a restricted Dec-POMDP model. We first transform the policy optimization problem to that of likelihood maximization in a mixture of dynamic Bayes nets (DBNs). We then develop the Expectation-Maximization (EM) algorithm for maximizing the likelihood in this representation. The EM algorithm for ND-POMDPs lends itself naturally to a simple message-passing paradigm guided by the agent interaction graph. It is thus highly scalable w.r.t. the number of agents, can be easily parallelized, and produces good quality solutions.
展开▼
