Approximate linear programming for constrained partially observable Markov decision processes
- Pascal Poupart, University of Waterloo
- Aarti Malhotra
- Pei Pei
- Kee-Eung Kim
- Bongseok Goh
- Michael Bowling, University of Alberta
In many situations, it is desirable to optimize a sequence of decisions by maximizing a primary objective while respecting some constraints with respect to secondary objectives. Such problems can be naturally modeled as constrained partially observable Markov decision processes (CPOMDPs) when the environment is partially observable. In this work, we describe a technique based on approximate linear programming to optimize policies in CPOMDPs. The optimization is performed offline and produces a finite state controller with desirable performance guarantees. The approach outperforms a constrained version of point-based value iteration on a suite of benchmark problems.
Citation
P. Poupart, A. Malhotra, P. Pei, K. Kim, B. Goh, M. Bowling. "Approximate linear programming for constrained partially observable Markov decision processes". National Conference on Artificial Intelligence (AAAI), (ed: Blai Bonet, Sven Koenig), pp 3342-3348, January 2015.| Keywords: | |
| Category: | In Conference |
| Web Links: | AAAI |
BibTeX
@incollection{Poupart+al:AAAI15,
author = {Pascal Poupart and Aarti Malhotra and Pei Pei and Kee-Eung Kim and
Bongseok Goh and Michael Bowling},
title = {Approximate linear programming for constrained partially observable
Markov decision processes},
Editor = {Blai Bonet, Sven Koenig},
Pages = {3342-3348},
booktitle = {National Conference on Artificial Intelligence (AAAI)},
year = 2015,
}Last Updated: October 29, 2020Submitted by Sabina P
