Sustaining behavioral diversity in NEAT

Hirotaka Moriguchi, Shinichi Honiden

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

7 Citations (Scopus)

Abstract

Niching schemes, which sustains population diversity and let an evolutionary population avoid premature convergence, have been extensively studied in the research field of evolutionary algorithms. Neuroevolutionary (NE) algorithms, such as NEAT, have also benefitted from niching. However, the latest research indicates that the use of genotypeor phenotype-similarity-based niching schemes in NE algorithms is not highly effective because these schemes have difficulty sustaining the behavioral diversity in the environment. In this paper, we propose a novel niching scheme that takes into consideration both the phenotypic and behavioral diversity, and then integrate it with NEAT. An experimental analysis revealed that the proposed algorithm outperforms the original NEAT for various problem settings. More interestingly, it performs especially well for problems with a high noise level and large state space. Since these features are common in problems to which NEAT is applied, the proposed algorithm should be effective in practice.

Original languageEnglish
Title of host publicationProceedings of the 12th Annual Genetic and Evolutionary Computation Conference, GECCO '10
Pages611-618
Number of pages8
DOIs
Publication statusPublished - 2010 Aug 27
Externally publishedYes
Event12th Annual Genetic and Evolutionary Computation Conference, GECCO-2010 - Portland, OR
Duration: 2010 Jul 72010 Jul 11

Other

Other12th Annual Genetic and Evolutionary Computation Conference, GECCO-2010
CityPortland, OR
Period10/7/710/7/11

Fingerprint

Niching
Population Diversity
Premature Convergence
Experimental Analysis
Phenotype
Evolutionary algorithms
Evolutionary Algorithms
State Space
Integrate

Keywords

  • Behavioral diversity
  • NEAT
  • Neuroevolution
  • Niching
  • Premature convergence

ASJC Scopus subject areas

  • Computational Theory and Mathematics
  • Theoretical Computer Science

Cite this

Moriguchi, H., & Honiden, S. (2010). Sustaining behavioral diversity in NEAT. In Proceedings of the 12th Annual Genetic and Evolutionary Computation Conference, GECCO '10 (pp. 611-618) https://doi.org/10.1145/1830483.1830595

Sustaining behavioral diversity in NEAT. / Moriguchi, Hirotaka; Honiden, Shinichi.

Proceedings of the 12th Annual Genetic and Evolutionary Computation Conference, GECCO '10. 2010. p. 611-618.

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

Moriguchi, H & Honiden, S 2010, Sustaining behavioral diversity in NEAT. in Proceedings of the 12th Annual Genetic and Evolutionary Computation Conference, GECCO '10. pp. 611-618, 12th Annual Genetic and Evolutionary Computation Conference, GECCO-2010, Portland, OR, 10/7/7. https://doi.org/10.1145/1830483.1830595
Moriguchi H, Honiden S. Sustaining behavioral diversity in NEAT. In Proceedings of the 12th Annual Genetic and Evolutionary Computation Conference, GECCO '10. 2010. p. 611-618 https://doi.org/10.1145/1830483.1830595
Moriguchi, Hirotaka ; Honiden, Shinichi. / Sustaining behavioral diversity in NEAT. Proceedings of the 12th Annual Genetic and Evolutionary Computation Conference, GECCO '10. 2010. pp. 611-618
@inproceedings{80260e0dfcb94c919ab58d21cba023cc,
title = "Sustaining behavioral diversity in NEAT",
abstract = "Niching schemes, which sustains population diversity and let an evolutionary population avoid premature convergence, have been extensively studied in the research field of evolutionary algorithms. Neuroevolutionary (NE) algorithms, such as NEAT, have also benefitted from niching. However, the latest research indicates that the use of genotypeor phenotype-similarity-based niching schemes in NE algorithms is not highly effective because these schemes have difficulty sustaining the behavioral diversity in the environment. In this paper, we propose a novel niching scheme that takes into consideration both the phenotypic and behavioral diversity, and then integrate it with NEAT. An experimental analysis revealed that the proposed algorithm outperforms the original NEAT for various problem settings. More interestingly, it performs especially well for problems with a high noise level and large state space. Since these features are common in problems to which NEAT is applied, the proposed algorithm should be effective in practice.",
keywords = "Behavioral diversity, NEAT, Neuroevolution, Niching, Premature convergence",
author = "Hirotaka Moriguchi and Shinichi Honiden",
year = "2010",
month = "8",
day = "27",
doi = "10.1145/1830483.1830595",
language = "English",
isbn = "9781450300728",
pages = "611--618",
booktitle = "Proceedings of the 12th Annual Genetic and Evolutionary Computation Conference, GECCO '10",

}

TY - GEN

T1 - Sustaining behavioral diversity in NEAT

AU - Moriguchi, Hirotaka

AU - Honiden, Shinichi

PY - 2010/8/27

Y1 - 2010/8/27

N2 - Niching schemes, which sustains population diversity and let an evolutionary population avoid premature convergence, have been extensively studied in the research field of evolutionary algorithms. Neuroevolutionary (NE) algorithms, such as NEAT, have also benefitted from niching. However, the latest research indicates that the use of genotypeor phenotype-similarity-based niching schemes in NE algorithms is not highly effective because these schemes have difficulty sustaining the behavioral diversity in the environment. In this paper, we propose a novel niching scheme that takes into consideration both the phenotypic and behavioral diversity, and then integrate it with NEAT. An experimental analysis revealed that the proposed algorithm outperforms the original NEAT for various problem settings. More interestingly, it performs especially well for problems with a high noise level and large state space. Since these features are common in problems to which NEAT is applied, the proposed algorithm should be effective in practice.

AB - Niching schemes, which sustains population diversity and let an evolutionary population avoid premature convergence, have been extensively studied in the research field of evolutionary algorithms. Neuroevolutionary (NE) algorithms, such as NEAT, have also benefitted from niching. However, the latest research indicates that the use of genotypeor phenotype-similarity-based niching schemes in NE algorithms is not highly effective because these schemes have difficulty sustaining the behavioral diversity in the environment. In this paper, we propose a novel niching scheme that takes into consideration both the phenotypic and behavioral diversity, and then integrate it with NEAT. An experimental analysis revealed that the proposed algorithm outperforms the original NEAT for various problem settings. More interestingly, it performs especially well for problems with a high noise level and large state space. Since these features are common in problems to which NEAT is applied, the proposed algorithm should be effective in practice.

KW - Behavioral diversity

KW - NEAT

KW - Neuroevolution

KW - Niching

KW - Premature convergence

UR - http://www.scopus.com/inward/record.url?scp=77955900415&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77955900415&partnerID=8YFLogxK

U2 - 10.1145/1830483.1830595

DO - 10.1145/1830483.1830595

M3 - Conference contribution

SN - 9781450300728

SP - 611

EP - 618

BT - Proceedings of the 12th Annual Genetic and Evolutionary Computation Conference, GECCO '10

ER -