Approximate equivalence of markov decision processes

Eyal Even-Dar, Yishay Mansour

Research output: Contribution to journalConference articlepeer-review

Abstract

We consider the problem of finding the minimal ε-equivalent MDP for an MDP given in its tabular form. We show that the problem is NP-Hard and then give a bicriteria approximation algorithm to the problem. We suggest that the right measure for finding minimal ε-equivalent model is L1 rather than L∞ by giving both an example, which demonstrates the drawback of using L∞, and performance guarantees for using L1. In addition, we give a polynomial algorithm that decides whether two MDPs are equivalent.

Original languageEnglish
Pages (from-to)581-594
Number of pages14
JournalLecture Notes in Computer Science
Volume2777
DOIs
StatePublished - 2003
Event16th Annual Conference on Learning Theory and 7th Kernel Workshop, COLT/Kernel 2003 - Washington, DC, United States
Duration: 24 Aug 200327 Aug 2003

Fingerprint

Dive into the research topics of 'Approximate equivalence of markov decision processes'. Together they form a unique fingerprint.

Cite this