A co-adaptive approach to XCS classifier systems

Chung Yuan Huang*, Chuen Tsai Sun, Yen Wei Chu

*Corresponding author for this work

Research output: Contribution to journalJournal Article peer-review

Abstract

We propose a co-adaptive approach to control coevolution-based eXtended Classifier System (XCS) parameters. By taking advantage of the on-line incremental learning capability of such systems, solutions can be produced that completely cover a target problem. The system combines the advantages of both adaptive and self-adaptive parameter-control approaches. Using a coevolution model means that two XCS can operate in parallel to simultaneously solve target and parameter-setting problems. Furthermore, the approach needs very little time to become efficient in terms of latent learning, since it only requires small amounts of information on performance metrics during early run-time stages. Test results indicate that our proposed system outperforms comparable models regardless of the target problem's stationary/non-stationary status.

Original languageEnglish
Pages (from-to)388-394
Number of pages7
JournalWSEAS Transactions on Systems
Volume5
Issue number2
StatePublished - 02 2006
Externally publishedYes

Keywords

  • CA-XCS
  • Co-adaptive
  • Coevolution
  • Dyna
  • Parameter adaptation
  • Parameter setting problem

Fingerprint

Dive into the research topics of 'A co-adaptive approach to XCS classifier systems'. Together they form a unique fingerprint.

Cite this