Date of Award
Master of Science
This work details a machine learning tool developed to support computational, agent-based simulation research in the social sciences. Specifically, the Java Reinforcement Learning Module (JReLM) is a platform for implementing reinforcement learning algorithms for use in agent-based simulations. The module was designed for use with the Recursive Porous Agent Simulation Toolkit (Repast), an agent-based simulation platform popular in computational social science research. Background, architecture, and implementation of JReLM are discussed within. This includes explanation of pre-implemented tools and algorithms available for immediate use in Repast simulations. In addition, an account of JReLM's use in an agent-based computational economics simulation is included as an illustrative application. Directions for further development and future use in ongoing agent-based computational economics work are discussed as well.
Gieseler, Charles, "A Java Reinforcement Learning Module for the Recursive Porous Agent Simulation Toolkit : facilitating study and experimentation with reinforcement learning in social science multi-agent simulations" (2005). Retrospective Theses and Dissertations. 18812.