Using mathematical programming to solve Factored Markov Decision Processes with Imprecise Probabilities

Karina Valdivia Delgado, Leliane Nunes De Barros, Fabio Gagliardi Cozman, Scott Sanner

Research output: Contribution to journalArticlepeer-review

6 Scopus citations

Abstract

This paper investigates Factored Markov Decision Processes with Imprecise Probabilities (MDPIPs); that is, Factored Markov Decision Processes (MDPs) where transition probabilities are imprecisely specified. We derive efficient approximate solutions for Factored MDPIPs based on mathematical programming. To do this, we extend previous linear programming approaches for linear approximations in Factored MDPs, resulting in a multilinear formulation for robust "maximin" linear approximations in Factored MDPIPs. By exploiting the factored structure in MDPIPs we are able to demonstrate orders of magnitude reduction in solution time over standard exact non-factored approaches, in exchange for relatively low approximation errors, on a difficult class of benchmark problems with millions of states. © 2011 Elsevier Inc. All rights reserved.
Original languageAmerican English
Pages (from-to)1000-1017
Number of pages18
JournalInternational Journal of Approximate Reasoning
DOIs
StatePublished - 1 Oct 2011
Externally publishedYes

Fingerprint

Dive into the research topics of 'Using mathematical programming to solve Factored Markov Decision Processes with Imprecise Probabilities'. Together they form a unique fingerprint.

Cite this