Please use this identifier to cite or link to this item: https://doi.org/10.1016/j.ins.2011.01.001
Title: Hessian matrix distribution for Bayesian policy gradient reinforcement learning
Authors: Vien, N.A. 
Yu, H.
Chung, T.
Keywords: Bayesian policy gradient
Hessian matrix distribution
Markov decision process
Monte-Carlo policy gradient
Policy gradient
Reinforcement learning
Issue Date: 2011
Citation: Vien, N.A., Yu, H., Chung, T. (2011). Hessian matrix distribution for Bayesian policy gradient reinforcement learning. Information Sciences 181 (9) : 1671-1685. ScholarBank@NUS Repository. https://doi.org/10.1016/j.ins.2011.01.001
Abstract: Bayesian policy gradient algorithms have been recently proposed for modeling the policy gradient of the performance measure in reinforcement learning as a Gaussian process. These methods were known to reduce the variance and the number of samples needed to obtain accurate gradient estimates in comparison to the conventional Monte-Carlo policy gradient algorithms. In this paper, we propose an improvement over previous Bayesian frameworks for the policy gradient. We use the Hessian matrix distribution as a learning rate schedule to improve the performance of the Bayesian policy gradient algorithm in terms of the variance and the number of samples. As in computing the policy gradient distributions, the Bayesian quadrature method is used to estimate the Hessian matrix distributions. We prove that the posterior mean of the Hessian distribution estimate is symmetric, one of the important properties of the Hessian matrix. Moreover, we prove that with an appropriate choice of kernel, the computational complexity of Hessian distribution estimate is equal to that of the policy gradient distribution estimates. Using simulations, we show encouraging experimental results comparing the proposed algorithm to the Bayesian policy gradient and the Bayesian policy natural gradient algorithms described in Ghavamzadeh and Engel [10]. © 2011 Elsevier Inc. All rights reserved.
Source Title: Information Sciences
URI: http://scholarbank.nus.edu.sg/handle/10635/38871
ISSN: 00200255
DOI: 10.1016/j.ins.2011.01.001
Appears in Collections:Staff Publications

Show full item record
Files in This Item:
There are no files associated with this item.

SCOPUSTM   
Citations

22
checked on Aug 10, 2018

WEB OF SCIENCETM
Citations

19
checked on Jul 9, 2018

Page view(s)

84
checked on Jul 27, 2018

Google ScholarTM

Check

Altmetric


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.