Java Resources for Teaching Reinforcement Learning


Christopher J. La Pilla: Class of 2004

Michael D. Schompert: Class of 2003

Document Type

Conference Material

Publication Date


Department 1

Computer Science


In this paper we present a library of classes for programming reinforcement learning simulations in Java. This library is based upon the standard by Sutton and Santamaria, with valuable additions to simplify the implementation of the software for selected temporal-difference control algorithms and various memory models. We also present arguments for the integration of this library into the curriculum of a Java-based undergraduate course in Artificial Intelligence.