Self organizing classifiers and niched fitness

Research output: Chapter in Book/Report/Conference proceedingConference contribution

11 Citations (Scopus)

Abstract

Learning classifier systems are adaptive learning systems which have been widely applied in a multitude of application domains. However, there are still some generalization problems unsolved. The hurdle is that fitness and niching pressures are difficult to balance. Here, a new algorithm called Self Organizing Classifiers is proposed which faces this problem from a different perspective. Instead of balancing the pressures, both pressures are separated and no balance is necessary. In fact, the proposed algorithm possesses a dynamical population structure that self-organizes itself to better project the input space into a map. The niched fitness concept is defined along with its dynamical population structure, both are indispensable for the understanding of the proposed method. Promising results are shown on two continuous multi-step problems. One of which is yet more challenging than previous problems of this class in the literature.

Original languageEnglish
Title of host publicationGECCO 2013 - Proceedings of the 2013 Genetic and Evolutionary Computation Conference
Pages1109-1116
Number of pages8
DOIs
Publication statusPublished - Sep 2 2013
Event2013 15th Genetic and Evolutionary Computation Conference, GECCO 2013 - Amsterdam, Netherlands
Duration: Jul 6 2013Jul 10 2013

Publication series

NameGECCO 2013 - Proceedings of the 2013 Genetic and Evolutionary Computation Conference

Other

Other2013 15th Genetic and Evolutionary Computation Conference, GECCO 2013
CountryNetherlands
CityAmsterdam
Period7/6/137/10/13

Fingerprint

Self-organizing
Fitness
Classifiers
Classifier
Pressure
Population Structure
Learning
Adaptive systems
Niching
Learning Classifier Systems
Population
Adaptive Learning
Learning systems
Adaptive Systems
Learning Systems
Balancing
Necessary

All Science Journal Classification (ASJC) codes

  • Genetics
  • Computational Mathematics

Cite this

Vargas, D. V., Takano, H., & Murata, J. (2013). Self organizing classifiers and niched fitness. In GECCO 2013 - Proceedings of the 2013 Genetic and Evolutionary Computation Conference (pp. 1109-1116). (GECCO 2013 - Proceedings of the 2013 Genetic and Evolutionary Computation Conference). https://doi.org/10.1145/2463372.2463501

Self organizing classifiers and niched fitness. / Vargas, Danilo V.; Takano, Hirotaka; Murata, Junichi.

GECCO 2013 - Proceedings of the 2013 Genetic and Evolutionary Computation Conference. 2013. p. 1109-1116 (GECCO 2013 - Proceedings of the 2013 Genetic and Evolutionary Computation Conference).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Vargas, DV, Takano, H & Murata, J 2013, Self organizing classifiers and niched fitness. in GECCO 2013 - Proceedings of the 2013 Genetic and Evolutionary Computation Conference. GECCO 2013 - Proceedings of the 2013 Genetic and Evolutionary Computation Conference, pp. 1109-1116, 2013 15th Genetic and Evolutionary Computation Conference, GECCO 2013, Amsterdam, Netherlands, 7/6/13. https://doi.org/10.1145/2463372.2463501
Vargas DV, Takano H, Murata J. Self organizing classifiers and niched fitness. In GECCO 2013 - Proceedings of the 2013 Genetic and Evolutionary Computation Conference. 2013. p. 1109-1116. (GECCO 2013 - Proceedings of the 2013 Genetic and Evolutionary Computation Conference). https://doi.org/10.1145/2463372.2463501
Vargas, Danilo V. ; Takano, Hirotaka ; Murata, Junichi. / Self organizing classifiers and niched fitness. GECCO 2013 - Proceedings of the 2013 Genetic and Evolutionary Computation Conference. 2013. pp. 1109-1116 (GECCO 2013 - Proceedings of the 2013 Genetic and Evolutionary Computation Conference).
@inproceedings{dade039cfc524eb7975a12ed69192e36,
title = "Self organizing classifiers and niched fitness",
abstract = "Learning classifier systems are adaptive learning systems which have been widely applied in a multitude of application domains. However, there are still some generalization problems unsolved. The hurdle is that fitness and niching pressures are difficult to balance. Here, a new algorithm called Self Organizing Classifiers is proposed which faces this problem from a different perspective. Instead of balancing the pressures, both pressures are separated and no balance is necessary. In fact, the proposed algorithm possesses a dynamical population structure that self-organizes itself to better project the input space into a map. The niched fitness concept is defined along with its dynamical population structure, both are indispensable for the understanding of the proposed method. Promising results are shown on two continuous multi-step problems. One of which is yet more challenging than previous problems of this class in the literature.",
author = "Vargas, {Danilo V.} and Hirotaka Takano and Junichi Murata",
year = "2013",
month = "9",
day = "2",
doi = "10.1145/2463372.2463501",
language = "English",
isbn = "9781450319638",
series = "GECCO 2013 - Proceedings of the 2013 Genetic and Evolutionary Computation Conference",
pages = "1109--1116",
booktitle = "GECCO 2013 - Proceedings of the 2013 Genetic and Evolutionary Computation Conference",

}

TY - GEN

T1 - Self organizing classifiers and niched fitness

AU - Vargas, Danilo V.

AU - Takano, Hirotaka

AU - Murata, Junichi

PY - 2013/9/2

Y1 - 2013/9/2

N2 - Learning classifier systems are adaptive learning systems which have been widely applied in a multitude of application domains. However, there are still some generalization problems unsolved. The hurdle is that fitness and niching pressures are difficult to balance. Here, a new algorithm called Self Organizing Classifiers is proposed which faces this problem from a different perspective. Instead of balancing the pressures, both pressures are separated and no balance is necessary. In fact, the proposed algorithm possesses a dynamical population structure that self-organizes itself to better project the input space into a map. The niched fitness concept is defined along with its dynamical population structure, both are indispensable for the understanding of the proposed method. Promising results are shown on two continuous multi-step problems. One of which is yet more challenging than previous problems of this class in the literature.

AB - Learning classifier systems are adaptive learning systems which have been widely applied in a multitude of application domains. However, there are still some generalization problems unsolved. The hurdle is that fitness and niching pressures are difficult to balance. Here, a new algorithm called Self Organizing Classifiers is proposed which faces this problem from a different perspective. Instead of balancing the pressures, both pressures are separated and no balance is necessary. In fact, the proposed algorithm possesses a dynamical population structure that self-organizes itself to better project the input space into a map. The niched fitness concept is defined along with its dynamical population structure, both are indispensable for the understanding of the proposed method. Promising results are shown on two continuous multi-step problems. One of which is yet more challenging than previous problems of this class in the literature.

UR - http://www.scopus.com/inward/record.url?scp=84883056658&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84883056658&partnerID=8YFLogxK

U2 - 10.1145/2463372.2463501

DO - 10.1145/2463372.2463501

M3 - Conference contribution

AN - SCOPUS:84883056658

SN - 9781450319638

T3 - GECCO 2013 - Proceedings of the 2013 Genetic and Evolutionary Computation Conference

SP - 1109

EP - 1116

BT - GECCO 2013 - Proceedings of the 2013 Genetic and Evolutionary Computation Conference

ER -