/rkhs

Importance Sampling Policy Gradient Algorithms in Reproducing Kernel Hilbert Space

Primary LanguageMatlab

rkhs

Importance Sampling Policy Gradient Algorithms in Reproducing Kernel Hilbert Space

Abstract

Modeling policies in Reproducing Kernel Hilbert Space (RKHS) offers a very flexible and powerful new family of policy gradient algorithms called RKHS policy gradient algorithms. They are designed to optimize over a space of very high or infinite dimensional policies. As a matter of fact, they are known to suffer from a large variance problem. This critical issue comes from the fact that updating the current policy is based on a functional gradient that does not exploit all old episodes sampled by previous policies. In this paper, we introduce a generalized RKHS policy gradient algorithm that integrates the following important ideas: i) policy modeling in RKHS; ii) normalized importance sampling, which helps reduce the estimation variance by reusing previously sampled episodes in a principled way; and iii) regularization terms, which avoid updating the policy too over-fit to sampled data. In the experiment section, we provide an analysis of the proposed algorithms through bench-marking domains. The experiment results show that the proposed algorithm can still enjoy a powerful policy modeling in RKHS and achieve more data-efficiency.

Code

  • Matlab 2014b

Paper

https://link.springer.com/article/10.1007/s10462-017-9579-x