MILCS: a mutual information learning classifier system

This paper introduces a new variety of learning classifier system (LCS), called MILCS, which utilizes mutual information as fitness feedback. Unlike most LCSs, MILCS is specifically designed for supervised learning. MILCS’s design draws on an analogy to the structural learning approach of cascade correlation networks. We present preliminary results, and contrast them to results from XCS. We discuss the explanatory power of the resulting rule sets, and introduce a new technique for visualizing explanatory power. Final comments include future directions for this research, including investigations in neural networks and other systems. Categories and Subject Descriptors I.2.6 [Artificial Intelligence]: Learning: Induction General Terms Algorithms Keywords Learning classifier systems, evolutionary computation, structural learning, supervised learning, cascade correlation, information theory, mutual information, visualization, explanatory power, rule learning.

• Type: Conference
• Year: 2007
• Where: GECCO
• Authors: Robert Elliott Smith, Max Kun Jiang
• Keywords: Explanatory Power, GECCO 2007, Learning Classifier Systems, Supervised Learning