Incremental Estimation of Natural Policy Gradient with Relative Importance Weighting

Ryo IWAKI  Hiroki YOKOYAMA  Minoru ASADA  

Publication
IEICE TRANSACTIONS on Information and Systems   Vol.E101-D   No.9   pp.2346-2355
Publication Date: 2018/09/01
Online ISSN: 1745-1361
DOI: 10.1587/transinf.2017EDP7363
Type of Manuscript: PAPER
Category: Artificial Intelligence, Data Mining
Keyword: 
reinforcement learning,  natural policy gradient,  adaptive step size,  

Full Text: PDF(3.1MB)
>>Buy this Article


Summary: 
The step size is a parameter of fundamental importance in learning algorithms, particularly for the natural policy gradient (NPG) methods. We derive an upper bound for the step size in an incremental NPG estimation, and propose an adaptive step size to implement the derived upper bound. The proposed adaptive step size guarantees that an updated parameter does not overshoot the target, which is achieved by weighting the learning samples according to their relative importances. We also provide tight upper and lower bounds for the step size, though they are not suitable for the incremental learning. We confirm the usefulness of the proposed step size using the classical benchmarks. To the best of our knowledge, this is the first adaptive step size method for NPG estimation.