Problem solution sustenance in XCS: Markov chain analysis of niche support distributions and the impact on computational complexity

Michigan-style learning classifier systems iteratively evolve a distributed solution to a problem in the form of potentially overlapping subsolutions. Each problem niche is covered by subsolutions that are represented by a set of predictive rules, termed classifiers. The genetic algorithm is designed to evolve classifier structures that together cover the whole problem space and represent a complete problem solution. An obvious challenge for such an online evolving, distributed knowledge representation is to continuously sustain all problem subsolutions covering all problem niches, that is, to ensure niche support. Effective niche support depends both on the probability of reproduction and on the probability of deletion of classifiers in a niche. In XCS, reproduction is occurrence-based whereas deletion is support-based. In combination, niche support is assured effectively. In this paper we present a Markov chain analysis of the niche support in XCS, which we validate experimentally. Evaluations in diverse Boolean function settings, which require non-overlapping and overlapping solution structures, support the theoretical derivations. We also consider the effects of mutation and crossover on niche support. With respect to computational complexity, the paper shows that XCS is able to maintain (partially overlapping) niches with a computational effort that is linear in the inverse of the niche occurrence frequency.

[1]  Martin V. Butz,et al.  Rule-Based Evolutionary Online Learning Systems - A Principled Approach to LCS Analysis and Design , 2006, Studies in Fuzziness and Soft Computing.

[2]  David E. Goldberg,et al.  The Gambler's Ruin Problem, Genetic Algorithms, and the Sizing of Populations , 1999, Evolutionary Computation.

[3]  Kenneth Alan De Jong,et al.  An analysis of the behavior of a class of genetic adaptive systems. , 1975 .

[4]  John H. Holland,et al.  Hidden Order: How Adaptation Builds Complexity , 1995 .

[5]  Martin J. Oates,et al.  A Preliminary Investigation of Modified XCS as a Generic Data Mining Tool , 2001, IWLCS.

[6]  Larry Bull,et al.  Self-Adaptive Mutation in ZCS Controllers , 2000, EvoWorkshops.

[7]  John R. Koza,et al.  Hidden Order: How Adaptation Builds Complexity. , 1995, Artificial Life.

[8]  Stewart W. Wilson Classifier Fitness Based on Accuracy , 1995, Evolutionary Computation.

[9]  Daniele Loiacono,et al.  Extending XCSF beyond linear approximation , 2005, GECCO '05.

[10]  F. Fairman Introduction to dynamic systems: Theory, models and applications , 1979, Proceedings of the IEEE.

[11]  Samir W. Mahfoud Crowding and Preselection Revisited , 1992, PPSN.

[12]  Kenneth A. De Jong,et al.  Learning Concept Classification Rules Using Genetic Algorithms , 1991, IJCAI.

[13]  Stewart W. Wilson ZCS: A Zeroth Level Classifier System , 1994, Evolutionary Computation.

[14]  David E. Goldberg,et al.  An Analysis of Reproduction and Crossover in a Binary-Coded Genetic Algorithm , 1987, ICGA.

[15]  T. Kovacs XCS Classifier System Reliably Evolves Accurate, Complete, and Minimal Representations for Boolean Functions , 1998 .

[16]  Martin V. Butz,et al.  An algorithmic description of XCS , 2000, Soft Comput..

[17]  K. Dejong,et al.  An analysis of the behavior of a class of genetic adaptive systems , 1975 .

[18]  M. Kendall Statistical Methods for Research Workers , 1937, Nature.

[19]  D. Goldberg,et al.  Bounding the Population Size to Ensure Niche Support in XCS , 2004 .

[20]  Thomas Stützle,et al.  GECCO 2007: Genetic and Evolutionary Computation Conference , 2007 .

[21]  D. Goldberg,et al.  Bounding Learning Time in XCS , 2004, GECCO.

[22]  Gilles Venturini,et al.  Adaptation in dynamic environments through a minimal probability of exploration , 1994 .

[23]  Georges R. Harik,et al.  Finding Multimodal Solutions Using Restricted Tournament Selection , 1995, ICGA.

[24]  David E. Goldberg,et al.  The gambler''s ruin problem , 1997 .

[25]  Stewart W. Wilson,et al.  Advances in learning classifier systems : Third International Workshop, IWLCS 2000, Paris, France, September 15-16, 2000 : revised papers , 2001 .

[26]  Tim Kovacs Strength or accuracy: credit assignment in learning classifier systems , 2003 .

[27]  D. E. Goldberg,et al.  Simple Genetic Algorithms and the Minimal, Deceptive Problem , 1987 .

[28]  Stewart W. Wilson,et al.  Advances in learning classifier systems : 4th International Workshop, IWLCS 2001, San Francisco, CA, USA, July 7-8, 2001 : revised papers , 2002 .

[29]  David J. Groggel,et al.  Practical Nonparametric Statistics , 2000, Technometrics.

[30]  J. David Schaffer,et al.  Proceedings of the third international conference on Genetic algorithms , 1989 .

[31]  Morgan B Kaufmann,et al.  Finite Markov Chain Analysis of Genetic Algorithms with Niching , 1993 .

[32]  Martin V. Butz,et al.  Strong, Stable, and Reliable Fitness Pressure in XCS due to Tournament Selection , 2005, Genetic Programming and Evolvable Machines.

[33]  Larry Bull,et al.  Simple Markov Models of the Genetic Algorithm in Classifier Systems: Accuracy-Based Fitness , 2000, IWLCS.

[34]  Martin V. Butz Kernel-based, ellipsoidal conditions in the real-valued XCS classifier system , 2005, GECCO '05.

[35]  Larry Bull,et al.  Self-adaptive mutation in classifier system controllers , 2000 .

[36]  David E. Goldberg,et al.  Finite Markov Chain Analysis of Genetic Algorithms , 1987, ICGA.

[37]  David E. Goldberg,et al.  Implicit Niching in a Learning Classifier System: Nature's Way , 1994, Evolutionary Computation.

[38]  John H. Holland,et al.  Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence , 1992 .

[39]  Ester Bernadó-Mansilla,et al.  Accuracy-Based Learning Classifier Systems: Models, Analysis and Applications to Classification Tasks , 2003, Evolutionary Computation.

[40]  T. Kovacs Deletion schemes for classifier systems , 1999 .

[41]  Michael Friendly,et al.  Visualizing Categorical Data , 2009, Encyclopedia of Database Systems.

[42]  J. Davenport Editor , 1960 .

[43]  Leonard Kleinrock,et al.  Theory, Volume 1, Queueing Systems , 1975 .

[44]  Stewart W. Wilson Generalization in the XCS Classifier System , 1998 .

[45]  Larry Bull,et al.  ZCS Redux , 2002, Evolutionary Computation.

[46]  Martin V. Butz,et al.  Analysis and Improvement of Fitness Exploitation in XCS: Bounding Models, Tournament Selection, and Bilateral Accuracy , 2003, Evolutionary Computation.

[47]  Larry Bull,et al.  A Self-Adaptive XCS , 2001, IWLCS.

[48]  Xavier Llorà,et al.  XCS and GALE: A Comparative Study of Two Learning Classifier Systems on Data Mining , 2001, IWLCS.

[49]  Stewart W. Wilson Classifier Systems and the Animat Problem , 1987, Machine Learning.

[50]  Stewart W. Wilson,et al.  Advances in Learning Classifier Systems. Fourth International Workshop , 2004 .

[51]  Martin V. Butz,et al.  Toward a theory of generalization and learning in XCS , 2004, IEEE Transactions on Evolutionary Computation.

[52]  Larry Bull,et al.  Simple Markov Models of the Genetic Algorithm in Classifier Systems: Multi-step Tasks , 2000, IWLCS.

[53]  Pier Luca Lanzi,et al.  An Analysis of Generalization in the XCS Classifier System , 1999, Evolutionary Computation.