Multiagent Stochastic Planning With Bayesian Policy Recognition