Active reinforcement learning

Arkady Epshteyn, Adam Vogel, Gerald Dejong

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

When the transition probabilities and rewards of a Markov Decision Process (MDP) are known, an agent can obtain the optimal policy without any interaction with the environment. However, exact transition probabilities are difficult for experts to specify. One option left to an agent is a long and potentially costly exploration of the environment. In this paper, we propose another alternative: given initial (possibly inaccurate) specification of the MDP, the agent determines the sensitivity of the optimal policy to changes in transitions and rewards. It then focuses its exploration on the regions of space to which the optimal policy is most sensitive. We show that the proposed exploration strategy performs well on several control and planning problems.

Original languageEnglish (US)
Title of host publicationProceedings of the 25th International Conference on Machine Learning
PublisherAssociation for Computing Machinery (ACM)
Pages296-303
Number of pages8
ISBN (Print)9781605582054
DOIs
StatePublished - Jan 1 2008
Event25th International Conference on Machine Learning - Helsinki, Finland
Duration: Jul 5 2008Jul 9 2008

Publication series

NameProceedings of the 25th International Conference on Machine Learning

Other

Other25th International Conference on Machine Learning
CountryFinland
CityHelsinki
Period7/5/087/9/08

ASJC Scopus subject areas

  • Artificial Intelligence
  • Human-Computer Interaction
  • Software

Fingerprint Dive into the research topics of 'Active reinforcement learning'. Together they form a unique fingerprint.

Cite this