%0 Conference Paper %1 wilson:1998:gXCScs %A Wilson, Stewart W. %B Genetic Programming 1998: Proceedings of the Third Annual Conference %C University of Wisconsin, Madison, Wisconsin, USA %D 1998 %E Koza, John R. %E Banzhaf, Wolfgang %E Chellapilla, Kumar %E Deb, Kalyanmoy %E Dorigo, Marco %E Fogel, David B. %E Garzon, Max H. %E Goldberg, David E. %E Iba, Hitoshi %E Riolo, Rick %I Morgan Kaufmann %K algorithms, classifiers genetic %P 665--674 %T Generalization in the XCS Classifier System %U http://citeseer.ist.psu.edu/148764.html %X This paper studies two changes to XCS, a classifier system in which fitness is based on prediction accuracy and the genetic algorithm takes place in environmental niches. The changes were aimed at increasing XCS's tendency to evolve accurate, maximally general classifiers and were tested on previously employed "woods" and multiplexer tasks. Together the changes bring XCS close to evolving populations whose high-fitness classifiers form a near-minimal, accurate, maximally general cover of the input and action product space. In addition, results on the multiplexer, a difficult categorization task, suggest that XCS's learning complexity is polynomial in the input length and thus may avoid the "curse of dimensionality", a notorious barrier to scale-up. A comparison between XCS and genetic programming in solving the 6multiplexer suggests that XCS's learning rate is about three orders of magnitude faster in terms of the number of input instances processed. %@ 1-55860-548-7

@inproceedings{wilson:1998:gXCScs, abstract = {This paper studies two changes to XCS, a classifier system in which fitness is based on prediction accuracy and the genetic algorithm takes place in environmental niches. The changes were aimed at increasing XCS's tendency to evolve accurate, maximally general classifiers and were tested on previously employed {"}woods{"} and multiplexer tasks. Together the changes bring XCS close to evolving populations whose high-fitness classifiers form a near-minimal, accurate, maximally general cover of the input and action product space. In addition, results on the multiplexer, a difficult categorization task, suggest that XCS's learning complexity is polynomial in the input length and thus may avoid the {"}curse of dimensionality{"}, a notorious barrier to scale-up. A comparison between XCS and genetic programming in solving the 6multiplexer suggests that XCS's learning rate is about three orders of magnitude faster in terms of the number of input instances processed.}, added-at = {2008-06-19T17:46:40.000+0200}, address = {University of Wisconsin, Madison, Wisconsin, USA}, author = {Wilson, Stewart W.}, biburl = {https://www.bibsonomy.org/bibtex/2a78f8027c99653c080eac7050aeb3cf1/brazovayeye}, booktitle = {Genetic Programming 1998: Proceedings of the Third Annual Conference}, editor = {Koza, John R. and Banzhaf, Wolfgang and Chellapilla, Kumar and Deb, Kalyanmoy and Dorigo, Marco and Fogel, David B. and Garzon, Max H. and Goldberg, David E. and Iba, Hitoshi and Riolo, Rick}, interhash = {3d4efb2a064f71a5d5acd2bae07fe5a0}, intrahash = {a78f8027c99653c080eac7050aeb3cf1}, isbn = {1-55860-548-7}, keywords = {algorithms, classifiers genetic}, month = {22-25 July}, notes = {GP-98}, pages = {665--674}, publisher = {Morgan Kaufmann}, publisher_address = {San Francisco, CA, USA}, timestamp = {2008-06-19T17:54:15.000+0200}, title = {Generalization in the {XCS} Classifier System}, url = {http://citeseer.ist.psu.edu/148764.html}, year = 1998 }