An Evolutionary Multiobjective Optimization-Based Learning Classifier System in Non-Markov Environment(<Special Issue>Information Systems and Human Sciences)

Abstract

Learning Classifier Systems (LCSs) are rule-based systems that automatically build their rule set so as to get optimal policies through evolutionary processes. This paper considers an evolutionary multiobjective optimization-based method for constructing LCSs that adjust to non-Markov environments. Our goal is to construct an XCSMH (eXtended Classifier System - Memory Hierarchic) that can obtain not only optimal policies but also highly generalized rule sets. Results of numerical experiments show that the proposed method is superior to an existing method with respect to the generality of the obtained rule sets.