Covariant Policy Search

August 2003

Covariant Policy Search


Abstract:

We investigate the problem of non-covariant behavior of policy gradient reinforcement learning algorithms. The policy gradient approach is amenable to analysis by information geometric methods. This leads us to propose a natural metric on controller parameterization that results from considering the manifold of probability distributions over paths induced by a stochastic controller. Investigation of this approach leads to a covariant gradient ascent rule. Interesting properties of this rule are discussed, including its relation with actor-critic style reinforcement learning algorithms. The algorithms discussed here are computationally quite efficient and on some interesting problems lead to dramatic performance improvement over non-covariant rules.
@conference{Bagnell-2003-8727,
author = {J. Andrew (Drew) Bagnell And Jeff Schneider},
title = {Covariant Policy Search},
booktitle = {Proceedings of 18th International Joint Conference on Artificial Intelligence (IJCAI '03)},
year = {2003},
month = {August},
pages = {1019 - 1024},
keywords = {Reinforcement Learning, Reinforce, Natural gradient, covariant},
}
Copyright notice: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. These works may not be reposted without the explicit permission of the copyright holder.