Report on the 2008 Reinforcement Learning Competition

  • Shimon Whiteson University of Amsterdam
  • Brian Tanner University of Alberta
  • Adam White University of Alberta
Keywords: reinforcement learning

Abstract

This article reports on the 2008 Reinforcement Learning Competition,
  which began in November 2007 and ended with a workshop at the
  International Conference on Machine Learning (ICML) in July 2008 in
  Helsinki, Finland.  Researchers from around the world developed
  reinforcement learning agents to compete in six problems of various
  complexity and difficulty.  The competition employed fundamentally
  redesigned evaluation frameworks that, unlike those in previous
  competitions, aimed to systematically encourage the submission of
  robust learning methods. We describe the unique challenges of
  empirical evaluation in reinforcement learning and briefly review
  the history of the previous competitions and the evaluation
  frameworks they employed.  We also describe the novel frameworks
  developed for the 2008 competition as well as the software
  infrastructure on which they rely.  Furthermore, we describe the six
  competition domains and present a summary of selected competition
  results.  Finally, we discuss the implications of these results and
  outline ideas for the future of the competition.

Author Biographies

Shimon Whiteson, University of Amsterdam
Shimon Whiteson is an assistant professor at the Instituut voor Infomatica at the Universiteit van Amsterdam.  His research focuses on reinforcement learning, evolutionary computation, and multiagent
systems.  He served as the chair of the organizing committee for the 2008 Reinforcement Learning Competition.
Brian Tanner, University of Alberta
Brian Tanner is a provisional Ph.D candidate at the University of Alberta.  His research focuses on empirical evaluation and comparison of reinforcement learning algorithms.  He served as the chair of the technical committee for the 2008 Reinforcement Learning Competition.
Adam White, University of Alberta
Adam White is a provisional Ph.D candidate at the University of Alberta.  His research focuses human-computer interaction using reinforcement learning.  He served as a member of the organizing committee for the 2008 Reinforcement Learning Competition.

Published
2010-06-28
Section
Articles