Controlled Markov Decision Processes with AVaR criteria for unbounded costs

JOURNAL OF COMPUTATIONAL AND APPLIED MATHEMATICS

卷:319

Article

Ugurlu, Kerem¹ 

[1] Univ Washington, Dept Appl Math, Seattle, WA 98195 USA

关键词: Markov decision problem;    Average-Value-at-Risk;    Optimal control;   

DOI  :  10.1016/j.cam.2016.11.052

来源: Elsevier

【 摘 要 】

In this paper, we consider the control problem with the Average-Value-at-Risk (AVaR) criteria of the possibly unbounded L-1-costs in infinite horizon on a Markov Decision Process (MDP). With a suitable state aggregation and by choosing a priori a global variables heuristically, we show that there exist optimal policies for the infinite horizon problem for possibly unbounded costs. (C) 2016 Elsevier B.V. All rights reserved.

【 授权许可】

Free