Predictive Representations for Policy Gradient in POMDPs
Published on Aug 26, 20093747 Views
We consider the problem of estimating the policy gradient in Partially Observable Markov Decision Processes (POMDPs) with a special class of policies that are based on Predictive State Representations