%0 Report %T Analysis of a Classification-based Policy Iteration Algorithm %+ Sequential Learning (SEQUEL) %A Lazaric, Alessandro %A Ghavamzadeh, Mohammad %A Munos, Remi %8 2010-05-07 %D 2010 %Z Cognitive science/Computer scienceReports %X We present a classification-based policy iteration algorithm, called Direct Policy Iteration, and provide its finite-sample analysis. Our results state a performance bound in terms of the number of policy improvement steps, the number of rollouts used in each iteration, the capacity of the considered policy space, and a new capacity measure which indicates how well the policy space can approximate policies that are greedy w.r.t. any of its members. The analysis reveals a tradeoff between the estimation and approximation errors in this classification-based policy iteration setting. We also study the consistency of the method when there exists a sequence of policy spaces with increasing capacity. %G English %2 https://inria.hal.science/inria-00482065v2/document %2 https://inria.hal.science/inria-00482065v2/file/dpi-jmlr.pdf %L inria-00482065 %U https://inria.hal.science/inria-00482065