Robust Quadratic Programming for MDPs with uncertain observation noise

The problem of Markov decision processes (MDPs) with uncertain observation noise has rarely been studied. This paper proposes a Robust Quadratic Programming (RQP) approach to approximate Bellman equation solution. Besides efficiency, the proposed algorithm exhibits great robustness against uncertain...

Full description

Saved in:
Bibliographic Details
Published in:Neurocomputing (Amsterdam) 2019-12, Vol.370, p.28-38
Main Authors: Su, Jianmei, Cheng, Hong, Guo, Hongliang, Peng, Zhinan
Format: Article
Language:English
Subjects:
Bellman equation solution
Different kernel forms
Robust quadratic programming
Uncertain observation noise
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The problem of Markov decision processes (MDPs) with uncertain observation noise has rarely been studied. This paper proposes a Robust Quadratic Programming (RQP) approach to approximate Bellman equation solution. Besides efficiency, the proposed algorithm exhibits great robustness against uncertain observation noise, which is essential in real world applications. We further represent the solution into kernel forms, which implicitly expands the state-encoded feature space to higher or even infinite dimensions. Experimental results well justify its efficiency and robustness. The comparison with different kernels demonstrates its flexibility of kernel selection for different application scenarios.
ISSN:0925-2312
1872-8286
DOI:10.1016/j.neucom.2019.08.045