Analysis and improvement of the genetic discovery component of XCS
作者： Sergio Morales-Ortigosa , Albert Orriols-Puig , Ester Bernadó-Mansilla
关键词:
摘要: XCS is a learning classifier system that uses genetic algorithms to evolve population of classifiers online. When applied classification problems described by continuous attributes, has demonstrated be able models - represented as set independent interval-based rules are, at least, accurate those created some the most competitive machine techniques such C4.5. Despite these successful results, analyses how different operators affect rule evolution for representation are lacking. This paper focuses on this issue and conducts systematic experimental analysis effect operators. The observations conclusions drawn from used tool designing new enable extract more than obtained original scheme. More specifically, provided with discovery component based strategies, crossover operator designed both one strategies. In all cases, behavior carefully analyzed compared ones XCS. overall enables us supply important insights into improve in real-world domains average.
medra.org 参考文章(32)
Tim Kovacs, Deletion schemes for classifier systems genetic and evolutionary computation conference. pp. 329- 336 ,(1999)
John H. Holland, Adaptation in natural and artificial systems ,(1992)
Ester Bernadó, Xavier Llorà, Josep M. Garrell, XCS and GALE: A Comparative Study of Two Learning Classifier Systems on Data Mining IWLCS '01 Revised Papers from the 4th International Workshop on Advances in Learning Classifier Systems. pp. 115- 132 ,(2001) , 10.1007/3-540-48104-4_8
Ingo Rechenberg, Evolutionsstrategie : Optimierung technischer Systeme nach Prinzipien der biologischen Evolution Frommann-Holzboog. ,(1973)
Damien Ernst, Arthur Louette, Introduction to Reinforcement Learning MIT Press. ,(1998)
F. Herrera, M. Lozano, J.L. Verdegay, Tackling Real-Coded Genetic Algorithms: Operators and Tools for Behavioural Analysis Artificial Intelligence Review. ,vol. 12, pp. 265- 319 ,(1998) , 10.1023/A:1006504901164
Janez Demšar, Statistical Comparisons of Classifiers over Multiple Data Sets Journal of Machine Learning Research. ,vol. 7, pp. 1- 30 ,(2006)
Frank Wilcoxon, Individual Comparisons by Ranking Methods Springer Series in Statistics. ,vol. 1, pp. 196- 202 ,(1992) , 10.1007/978-1-4612-4380-9_16
Milton Friedman, The Use of Ranks to Avoid the Assumption of Normality Implicit in the Analysis of Variance Journal of the American Statistical Association. ,vol. 32, pp. 675- 701 ,(1937) , 10.2307/2279372