Many empirical results in reinforcement learning are based on a very small set of environments. These results often represent the best algorithm parameters that were found after an ad-hoc tuning or fitting process. We argue that presenting tuned scores from a small set of environments leads to method overfitting, wherein results may not generalize to similar environments. To address this problem, we advocate empirical evaluations using generalized domains: parameterized problem generators that explicitly encode variations in the environment to which the learner should be robust. We argue that evaluating across a set of these generated problems offers a more meaningful evaluation of reinforcement learning algorithms.

PDF, HTML

In ICML Workshop on Evaluation Methods for Machine Learning, June 2009. To appear..

@InProceedings{ICML09ws-shimon, title={Generalized Domains for Empirical Evaluations in Reinforcement Learning}, author={Shimon Whiteson and Brian Tanner and Matthew E. Taylor and Peter Stone}, booktitle={ICML Workshop on Evaluation Methods for Machine Learning}, month={June}, note={To appear.}, url="http://nn.cs.utexas.edu/?whiteson:icml09ws", year={2009} }