Abstract
This text presents a classifier system (CS), which is able to adapt to an environment by adjusting the activation probabilities of the rules and changing the rules itself. The operators for changing the rules are incorporated into the CS, thus allowing for an adaption of the rates of change on-line during the search process for better rules. An age is attached to the rules. Removal of rules from the rule set is done according to the age.
Experiments show that this approach to adapting the rule set by means of internal genetic operators (GO) is superior to exogenous genetic operators.
Preview
Unable to display preview. Download preview PDF.
References
John J. Grefenstette. Multilevel credit assignment in a genetic learning system. In John J. Grefenstette, editor, Genetic Algorithms and their Applications: Proceedings of the Second International Conference on Genetic Algorithms, pages 202–209, Lawrence Erlbaum Associates, 1987.
John H. Holland, Keith J. Holyhoak, Richard E. Nisbett, and Paul R. Thargard. Induction: Processes of Inference, Learning and Discovery. MIT Press, 1986.
John J. Holland. Adaption in natural and artificial systems. University of Michigan, 1975.
Andreas Schachtner. Problemlösen mittels einer broadcast language, 1989. Thesis, German.
J. David Schaffer and Amy Morishima. An adaptive crossover distribution mechanism for genetic algorithms. In John J. Grefenstette, editor, Genetic Algorithms and their Applications: Proceedings of the Second International Conference on Genetic Algorithms, pages 36–40, Lawrence Erlbaum Associates, 1987.
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1991 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Schachtner, A. (1991). A classifier system with integrated genetic operators. In: Schwefel, HP., Männer, R. (eds) Parallel Problem Solving from Nature. PPSN 1990. Lecture Notes in Computer Science, vol 496. Springer, Berlin, Heidelberg. https://doi.org/10.1007/BFb0029773
Download citation
DOI: https://doi.org/10.1007/BFb0029773
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-54148-6
Online ISBN: 978-3-540-70652-6
eBook Packages: Springer Book Archive