Action Elimination and Stopping Conditions for Reinforcement Learning

Eyal Even-Dar*, Shie Mannor, Yishay Mansour

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

28 Scopus citations

Abstract

We consider incorporating action elimination procedures in reinforcement learning algorithms. We suggest a framework that is based on learning an upper and a lower estimates of the value function or the Q-function and eliminating actions that are not optimal. We provide a model-based and a model-free variants of the elimination method. We further derive stopping conditions that guarantee that the learned policy is approximately optimal with high probability. Simulations demonstrate a considerable speedup and added robustness.

Original languageEnglish
Title of host publicationProceedings, Twentieth International Conference on Machine Learning
EditorsT. Fawcett, N. Mishra
Pages162-169
Number of pages8
StatePublished - 2003
EventProceedings, Twentieth International Conference on Machine Learning - Washington, DC, United States
Duration: 21 Aug 200324 Aug 2003

Publication series

NameProceedings, Twentieth International Conference on Machine Learning
Volume1

Conference

ConferenceProceedings, Twentieth International Conference on Machine Learning
Country/TerritoryUnited States
CityWashington, DC
Period21/08/0324/08/03

Fingerprint

Dive into the research topics of 'Action Elimination and Stopping Conditions for Reinforcement Learning'. Together they form a unique fingerprint.

Cite this