Sample-based Crowding method for multimodal optimization in continuous domain

Shin Ando, Einoshin Suzuki, Shigenobu Kobayashi

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

6 Citations (Scopus)

Abstract

We propose a selection scheme called Sample-based Crowding, which is aimed to improve the performance of Genetic Algorithms for multimodal optimization in ill-scaled and locally multimodal domains. These domains can be problematic for conventional approaches, but are commonly found in real-world optimization problems. The principle of Crowding is to apply a tournament selection to a parent-child pair with a high similarity. In the Sample-based Crowding, we determine such pairs based on a statistical comparison of the fitness values, which are sampled from the region between the pairs. Further, we take into account the ranks of the parents among the sampled values in the selection process, to determine their indispensability. These measurements are scale-invariant, which enables the proposed method to search a domain without presuming the distance between the optima or the scaling and the correlation of the variables. The proposed approach is evaluated in two bench-mark problems with an ill-scaled and a locally multi-modal landscape. The proposed method has a substantial advantage in terms of comprehensiveness compared to the conventional approaches, despite the additional cost of evaluations.

Original languageEnglish
Title of host publication2005 IEEE Congress on Evolutionary Computation, IEEE CEC 2005. Proceedings
Pages1867-1874
Number of pages8
Publication statusPublished - Oct 31 2005
Externally publishedYes
Event2005 IEEE Congress on Evolutionary Computation, IEEE CEC 2005 - Edinburgh, Scotland, United Kingdom
Duration: Sep 2 2005Sep 5 2005

Publication series

Name2005 IEEE Congress on Evolutionary Computation, IEEE CEC 2005. Proceedings
Volume2

Other

Other2005 IEEE Congress on Evolutionary Computation, IEEE CEC 2005
CountryUnited Kingdom
CityEdinburgh, Scotland
Period9/2/059/5/05

Fingerprint

Genetic algorithms
Costs

All Science Journal Classification (ASJC) codes

  • Engineering(all)

Cite this

Ando, S., Suzuki, E., & Kobayashi, S. (2005). Sample-based Crowding method for multimodal optimization in continuous domain. In 2005 IEEE Congress on Evolutionary Computation, IEEE CEC 2005. Proceedings (pp. 1867-1874). (2005 IEEE Congress on Evolutionary Computation, IEEE CEC 2005. Proceedings; Vol. 2).

Sample-based Crowding method for multimodal optimization in continuous domain. / Ando, Shin; Suzuki, Einoshin; Kobayashi, Shigenobu.

2005 IEEE Congress on Evolutionary Computation, IEEE CEC 2005. Proceedings. 2005. p. 1867-1874 (2005 IEEE Congress on Evolutionary Computation, IEEE CEC 2005. Proceedings; Vol. 2).

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

Ando, S, Suzuki, E & Kobayashi, S 2005, Sample-based Crowding method for multimodal optimization in continuous domain. in 2005 IEEE Congress on Evolutionary Computation, IEEE CEC 2005. Proceedings. 2005 IEEE Congress on Evolutionary Computation, IEEE CEC 2005. Proceedings, vol. 2, pp. 1867-1874, 2005 IEEE Congress on Evolutionary Computation, IEEE CEC 2005, Edinburgh, Scotland, United Kingdom, 9/2/05.
Ando S, Suzuki E, Kobayashi S. Sample-based Crowding method for multimodal optimization in continuous domain. In 2005 IEEE Congress on Evolutionary Computation, IEEE CEC 2005. Proceedings. 2005. p. 1867-1874. (2005 IEEE Congress on Evolutionary Computation, IEEE CEC 2005. Proceedings).
Ando, Shin ; Suzuki, Einoshin ; Kobayashi, Shigenobu. / Sample-based Crowding method for multimodal optimization in continuous domain. 2005 IEEE Congress on Evolutionary Computation, IEEE CEC 2005. Proceedings. 2005. pp. 1867-1874 (2005 IEEE Congress on Evolutionary Computation, IEEE CEC 2005. Proceedings).
@inproceedings{1aff181b81af49faa7fe18500d35fccd,
title = "Sample-based Crowding method for multimodal optimization in continuous domain",
abstract = "We propose a selection scheme called Sample-based Crowding, which is aimed to improve the performance of Genetic Algorithms for multimodal optimization in ill-scaled and locally multimodal domains. These domains can be problematic for conventional approaches, but are commonly found in real-world optimization problems. The principle of Crowding is to apply a tournament selection to a parent-child pair with a high similarity. In the Sample-based Crowding, we determine such pairs based on a statistical comparison of the fitness values, which are sampled from the region between the pairs. Further, we take into account the ranks of the parents among the sampled values in the selection process, to determine their indispensability. These measurements are scale-invariant, which enables the proposed method to search a domain without presuming the distance between the optima or the scaling and the correlation of the variables. The proposed approach is evaluated in two bench-mark problems with an ill-scaled and a locally multi-modal landscape. The proposed method has a substantial advantage in terms of comprehensiveness compared to the conventional approaches, despite the additional cost of evaluations.",
author = "Shin Ando and Einoshin Suzuki and Shigenobu Kobayashi",
year = "2005",
month = "10",
day = "31",
language = "English",
isbn = "0780393635",
series = "2005 IEEE Congress on Evolutionary Computation, IEEE CEC 2005. Proceedings",
pages = "1867--1874",
booktitle = "2005 IEEE Congress on Evolutionary Computation, IEEE CEC 2005. Proceedings",

}

TY - GEN

T1 - Sample-based Crowding method for multimodal optimization in continuous domain

AU - Ando, Shin

AU - Suzuki, Einoshin

AU - Kobayashi, Shigenobu

PY - 2005/10/31

Y1 - 2005/10/31

N2 - We propose a selection scheme called Sample-based Crowding, which is aimed to improve the performance of Genetic Algorithms for multimodal optimization in ill-scaled and locally multimodal domains. These domains can be problematic for conventional approaches, but are commonly found in real-world optimization problems. The principle of Crowding is to apply a tournament selection to a parent-child pair with a high similarity. In the Sample-based Crowding, we determine such pairs based on a statistical comparison of the fitness values, which are sampled from the region between the pairs. Further, we take into account the ranks of the parents among the sampled values in the selection process, to determine their indispensability. These measurements are scale-invariant, which enables the proposed method to search a domain without presuming the distance between the optima or the scaling and the correlation of the variables. The proposed approach is evaluated in two bench-mark problems with an ill-scaled and a locally multi-modal landscape. The proposed method has a substantial advantage in terms of comprehensiveness compared to the conventional approaches, despite the additional cost of evaluations.

AB - We propose a selection scheme called Sample-based Crowding, which is aimed to improve the performance of Genetic Algorithms for multimodal optimization in ill-scaled and locally multimodal domains. These domains can be problematic for conventional approaches, but are commonly found in real-world optimization problems. The principle of Crowding is to apply a tournament selection to a parent-child pair with a high similarity. In the Sample-based Crowding, we determine such pairs based on a statistical comparison of the fitness values, which are sampled from the region between the pairs. Further, we take into account the ranks of the parents among the sampled values in the selection process, to determine their indispensability. These measurements are scale-invariant, which enables the proposed method to search a domain without presuming the distance between the optima or the scaling and the correlation of the variables. The proposed approach is evaluated in two bench-mark problems with an ill-scaled and a locally multi-modal landscape. The proposed method has a substantial advantage in terms of comprehensiveness compared to the conventional approaches, despite the additional cost of evaluations.

UR - http://www.scopus.com/inward/record.url?scp=27144547959&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=27144547959&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:27144547959

SN - 0780393635

T3 - 2005 IEEE Congress on Evolutionary Computation, IEEE CEC 2005. Proceedings

SP - 1867

EP - 1874

BT - 2005 IEEE Congress on Evolutionary Computation, IEEE CEC 2005. Proceedings

ER -