Evolutionary Intelligence

, Volume 3, Issue 1, pp 31–50

A learning classifier system with mutual-information-based fitness

Authors

  • Robert Elliott Smith
    • Department of Computer ScienceUniversity College London
    • Department of Computer ScienceUniversity College London
  • Jaume Bacardit
    • School of Computer ScienceUniversity of Nottingham
  • Michael Stout
    • School of Computer ScienceUniversity of Nottingham
  • Natalio Krasnogor
    • School of Computer ScienceUniversity of Nottingham
  • Jonathan D. Hirst
    • School of ChemistryUniversity of Nottingham
Research Paper

DOI: 10.1007/s12065-010-0037-9

Cite this article as:
Smith, R.E., Jiang, M.K., Bacardit, J. et al. Evol. Intel. (2010) 3: 31. doi:10.1007/s12065-010-0037-9

Abstract

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. We present experimental results, and contrast them to results from XCS, UCS, GAssist, BioHEL, C4.5 and Naïve Bayes. We discuss the explanatory power of the resulting rule sets. MILCS is also shown to promote the discovery of default hierarchies, an important advantage of LCSs. Final comments include future directions for this research, including investigations in neural networks and other systems.

Keywords

Evolutionary computationLearning classifier systemsMachine learningInformation theoryMutual informationSupervised learningProtein structure predictionExplanatory power

Copyright information

© Springer-Verlag 2010