Sciweavers

Share
ML
2008
ACM

Learning near-optimal policies with Bellman-residual minimization based fitted policy iteration and a single sample path

8 years 2 months ago
Learning near-optimal policies with Bellman-residual minimization based fitted policy iteration and a single sample path
Abstract. We consider batch reinforcement learning problems in continuous space, expected total discounted-reward Markovian Decision Problems. As opposed to previous theoretical work, we consider the case when the training data consists of a single sample path (trajectory) of some behaviour policy. In particular, we do not assume access to a generative model of the environment. The algorithm studied is policy iteration where in successive iterations the Q-functions of the intermediate policies are obtained by means of minimizing a novel Bellman-residual type error. PACstyle polynomial bounds are derived on the number of samples needed to guarantee near-optimal performance where the bound depends on the mixing rate of the trajectory, the smoothness properties of the underlying Markovian Decision Problem, the approximation power and capacity of the function set used.
András Antos, Csaba Szepesvári, R&ea
Added 13 Dec 2010
Updated 13 Dec 2010
Type Journal
Year 2008
Where ML
Authors András Antos, Csaba Szepesvári, Rémi Munos
Comments (0)
books