Peer-to-peer evolutionary algorithms with adaptive autonomous selection

W.R.M.U.K. Wickramasinghe, M. van Steen, A.E. Eiben

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

30 Citations (Scopus)

Abstract

In this paper we describe and evaluate a fully distributed P2P evolutionary algorithm (EA) with adaptive autonomous selection. Autonomous selection means that decisions regarding survival and reproduction are taken by the individuals themselves independently, without any central control.This allows for a fully distributed EA, where not only reproduction (crossover and mutation) but also selection is performed at local level. An unwanted consequence of adding and removing individuals in a non-synchronized manner is that the population size gets out of control too. This problem is resolved by addingan adaptation mechanism allowing individuals to regulate their own selection pressure. The key tothis is a gossiping algorithm that enables individuals to maintain estimates on the size andthe fitness of the population. The algorithm is experimentally evaluated on a test problem to show the viability of the idea and to gain insight into the run-time dynamics of such an algorithm. The results convincingly demonstrate the feasibility of a fully decentralized EA in which the population size can be kept stable.

Original languageEnglish
Title of host publicationGECCO '07
Subtitle of host publicationProceedings of the 9th Annual Conference on Genetic and Evolutionary Computation
Place of PublicationNew York, NY
PublisherACM Publishing
Pages1460-1467
Number of pages8
ISBN (Print)978-1-59593-697-4
DOIs
Publication statusPublished - 27 Aug 2007
Externally publishedYes
Event9th Genetic and Evolutionary Computation Conference, GECCO 2007 - University College London, London, United Kingdom
Duration: 7 Jul 200711 Jul 2007
Conference number: 9

Conference

Conference9th Genetic and Evolutionary Computation Conference, GECCO 2007
Abbreviated titleGECCO 2007
CountryUnited Kingdom
CityLondon
Period7/07/0711/07/07

Fingerprint

Peer to Peer
Evolutionary algorithms
Evolutionary Algorithms
Parallel algorithms
Population Size
Gossiping
Distributed Algorithms
Viability
Fitness
Decentralized
Test Problems
Crossover
Mutation
Evaluate
Estimate
Demonstrate

Keywords

  • Autonomous selection
  • Distributed EA
  • Gossiping
  • Newscast protocol
  • Parameter adaptation

Cite this

Wickramasinghe, W. R. M. U. K., van Steen, M., & Eiben, A. E. (2007). Peer-to-peer evolutionary algorithms with adaptive autonomous selection. In GECCO '07: Proceedings of the 9th Annual Conference on Genetic and Evolutionary Computation (pp. 1460-1467). New York, NY: ACM Publishing. https://doi.org/10.1145/1276958.1277225
Wickramasinghe, W.R.M.U.K. ; van Steen, M. ; Eiben, A.E. / Peer-to-peer evolutionary algorithms with adaptive autonomous selection. GECCO '07: Proceedings of the 9th Annual Conference on Genetic and Evolutionary Computation. New York, NY : ACM Publishing, 2007. pp. 1460-1467
@inproceedings{2b1172d66d494b97b79fd9c1055a253b,
title = "Peer-to-peer evolutionary algorithms with adaptive autonomous selection",
abstract = "In this paper we describe and evaluate a fully distributed P2P evolutionary algorithm (EA) with adaptive autonomous selection. Autonomous selection means that decisions regarding survival and reproduction are taken by the individuals themselves independently, without any central control.This allows for a fully distributed EA, where not only reproduction (crossover and mutation) but also selection is performed at local level. An unwanted consequence of adding and removing individuals in a non-synchronized manner is that the population size gets out of control too. This problem is resolved by addingan adaptation mechanism allowing individuals to regulate their own selection pressure. The key tothis is a gossiping algorithm that enables individuals to maintain estimates on the size andthe fitness of the population. The algorithm is experimentally evaluated on a test problem to show the viability of the idea and to gain insight into the run-time dynamics of such an algorithm. The results convincingly demonstrate the feasibility of a fully decentralized EA in which the population size can be kept stable.",
keywords = "Autonomous selection, Distributed EA, Gossiping, Newscast protocol, Parameter adaptation",
author = "W.R.M.U.K. Wickramasinghe and {van Steen}, M. and A.E. Eiben",
year = "2007",
month = "8",
day = "27",
doi = "10.1145/1276958.1277225",
language = "English",
isbn = "978-1-59593-697-4",
pages = "1460--1467",
booktitle = "GECCO '07",
publisher = "ACM Publishing",

}

Wickramasinghe, WRMUK, van Steen, M & Eiben, AE 2007, Peer-to-peer evolutionary algorithms with adaptive autonomous selection. in GECCO '07: Proceedings of the 9th Annual Conference on Genetic and Evolutionary Computation. ACM Publishing, New York, NY, pp. 1460-1467, 9th Genetic and Evolutionary Computation Conference, GECCO 2007, London, United Kingdom, 7/07/07. https://doi.org/10.1145/1276958.1277225

Peer-to-peer evolutionary algorithms with adaptive autonomous selection. / Wickramasinghe, W.R.M.U.K.; van Steen, M.; Eiben, A.E.

GECCO '07: Proceedings of the 9th Annual Conference on Genetic and Evolutionary Computation. New York, NY : ACM Publishing, 2007. p. 1460-1467.

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

TY - GEN

T1 - Peer-to-peer evolutionary algorithms with adaptive autonomous selection

AU - Wickramasinghe, W.R.M.U.K.

AU - van Steen, M.

AU - Eiben, A.E.

PY - 2007/8/27

Y1 - 2007/8/27

N2 - In this paper we describe and evaluate a fully distributed P2P evolutionary algorithm (EA) with adaptive autonomous selection. Autonomous selection means that decisions regarding survival and reproduction are taken by the individuals themselves independently, without any central control.This allows for a fully distributed EA, where not only reproduction (crossover and mutation) but also selection is performed at local level. An unwanted consequence of adding and removing individuals in a non-synchronized manner is that the population size gets out of control too. This problem is resolved by addingan adaptation mechanism allowing individuals to regulate their own selection pressure. The key tothis is a gossiping algorithm that enables individuals to maintain estimates on the size andthe fitness of the population. The algorithm is experimentally evaluated on a test problem to show the viability of the idea and to gain insight into the run-time dynamics of such an algorithm. The results convincingly demonstrate the feasibility of a fully decentralized EA in which the population size can be kept stable.

AB - In this paper we describe and evaluate a fully distributed P2P evolutionary algorithm (EA) with adaptive autonomous selection. Autonomous selection means that decisions regarding survival and reproduction are taken by the individuals themselves independently, without any central control.This allows for a fully distributed EA, where not only reproduction (crossover and mutation) but also selection is performed at local level. An unwanted consequence of adding and removing individuals in a non-synchronized manner is that the population size gets out of control too. This problem is resolved by addingan adaptation mechanism allowing individuals to regulate their own selection pressure. The key tothis is a gossiping algorithm that enables individuals to maintain estimates on the size andthe fitness of the population. The algorithm is experimentally evaluated on a test problem to show the viability of the idea and to gain insight into the run-time dynamics of such an algorithm. The results convincingly demonstrate the feasibility of a fully decentralized EA in which the population size can be kept stable.

KW - Autonomous selection

KW - Distributed EA

KW - Gossiping

KW - Newscast protocol

KW - Parameter adaptation

UR - http://www.scopus.com/inward/record.url?scp=34548098403&partnerID=8YFLogxK

U2 - 10.1145/1276958.1277225

DO - 10.1145/1276958.1277225

M3 - Conference contribution

SN - 978-1-59593-697-4

SP - 1460

EP - 1467

BT - GECCO '07

PB - ACM Publishing

CY - New York, NY

ER -

Wickramasinghe WRMUK, van Steen M, Eiben AE. Peer-to-peer evolutionary algorithms with adaptive autonomous selection. In GECCO '07: Proceedings of the 9th Annual Conference on Genetic and Evolutionary Computation. New York, NY: ACM Publishing. 2007. p. 1460-1467 https://doi.org/10.1145/1276958.1277225