Intensification strategies for extremal optimisation

Marcus Randall, Andrew Lewis

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

3 Citations (Scopus)

Abstract

It is only relatively recently that extremal optimisation (EO) has been applied to combinatorial optimisation problems. As such, there have been only a few attempts to extend the paradigm to include standard search mechanisms that are routinely used by other techniques such as genetic algorithms, tabu search and ant colony optimisation. The key way to begin this process is to augment EO with attributes that it naturally lacks. While EO does not get confounded by local optima and is able to move through search space unencumbered, one of the major issues is to provide it with better search intensification strategies. In this paper, two strategies that compliment EO's mechanics are introduced and are used to augment an existing solver framework. Results, for single and population versions of the algorithm, demonstrate that intensification aids the performance of EO.

Original languageEnglish
Title of host publicationSimulated Evolution and Learning - 8th International Conference, SEAL 2010, Proceedings
Pages115-124
Number of pages10
Volume6457 LNCS
DOIs
Publication statusPublished - 2010
Event8th International Conference on Simulated Evolution and Learning, SEAL 2010 - Kanpur, India
Duration: 1 Dec 20104 Dec 2010

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume6457 LNCS
ISSN (Print)03029743
ISSN (Electronic)16113349

Conference

Conference8th International Conference on Simulated Evolution and Learning, SEAL 2010
CountryIndia
CityKanpur
Period1/12/104/12/10

Fingerprint

Extremal Optimization
Tabu search
Ant colony optimization
Combinatorial optimization
Search Strategy
Tabu Search
Combinatorial Optimization Problem
Search Space
Mechanics
Genetic algorithms
Paradigm
Attribute
Genetic Algorithm
Strategy
Demonstrate

Cite this

Randall, M., & Lewis, A. (2010). Intensification strategies for extremal optimisation. In Simulated Evolution and Learning - 8th International Conference, SEAL 2010, Proceedings (Vol. 6457 LNCS, pp. 115-124). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6457 LNCS). https://doi.org/10.1007/978-3-642-17298-4_12
Randall, Marcus ; Lewis, Andrew. / Intensification strategies for extremal optimisation. Simulated Evolution and Learning - 8th International Conference, SEAL 2010, Proceedings. Vol. 6457 LNCS 2010. pp. 115-124 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{07dc4ed300bb4fd3b275f8dcbdf70bd0,
title = "Intensification strategies for extremal optimisation",
abstract = "It is only relatively recently that extremal optimisation (EO) has been applied to combinatorial optimisation problems. As such, there have been only a few attempts to extend the paradigm to include standard search mechanisms that are routinely used by other techniques such as genetic algorithms, tabu search and ant colony optimisation. The key way to begin this process is to augment EO with attributes that it naturally lacks. While EO does not get confounded by local optima and is able to move through search space unencumbered, one of the major issues is to provide it with better search intensification strategies. In this paper, two strategies that compliment EO's mechanics are introduced and are used to augment an existing solver framework. Results, for single and population versions of the algorithm, demonstrate that intensification aids the performance of EO.",
author = "Marcus Randall and Andrew Lewis",
year = "2010",
doi = "10.1007/978-3-642-17298-4_12",
language = "English",
isbn = "3642172970",
volume = "6457 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "115--124",
booktitle = "Simulated Evolution and Learning - 8th International Conference, SEAL 2010, Proceedings",

}

Randall, M & Lewis, A 2010, Intensification strategies for extremal optimisation. in Simulated Evolution and Learning - 8th International Conference, SEAL 2010, Proceedings. vol. 6457 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 6457 LNCS, pp. 115-124, 8th International Conference on Simulated Evolution and Learning, SEAL 2010, Kanpur, India, 1/12/10. https://doi.org/10.1007/978-3-642-17298-4_12

Intensification strategies for extremal optimisation. / Randall, Marcus; Lewis, Andrew.

Simulated Evolution and Learning - 8th International Conference, SEAL 2010, Proceedings. Vol. 6457 LNCS 2010. p. 115-124 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6457 LNCS).

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

TY - GEN

T1 - Intensification strategies for extremal optimisation

AU - Randall, Marcus

AU - Lewis, Andrew

PY - 2010

Y1 - 2010

N2 - It is only relatively recently that extremal optimisation (EO) has been applied to combinatorial optimisation problems. As such, there have been only a few attempts to extend the paradigm to include standard search mechanisms that are routinely used by other techniques such as genetic algorithms, tabu search and ant colony optimisation. The key way to begin this process is to augment EO with attributes that it naturally lacks. While EO does not get confounded by local optima and is able to move through search space unencumbered, one of the major issues is to provide it with better search intensification strategies. In this paper, two strategies that compliment EO's mechanics are introduced and are used to augment an existing solver framework. Results, for single and population versions of the algorithm, demonstrate that intensification aids the performance of EO.

AB - It is only relatively recently that extremal optimisation (EO) has been applied to combinatorial optimisation problems. As such, there have been only a few attempts to extend the paradigm to include standard search mechanisms that are routinely used by other techniques such as genetic algorithms, tabu search and ant colony optimisation. The key way to begin this process is to augment EO with attributes that it naturally lacks. While EO does not get confounded by local optima and is able to move through search space unencumbered, one of the major issues is to provide it with better search intensification strategies. In this paper, two strategies that compliment EO's mechanics are introduced and are used to augment an existing solver framework. Results, for single and population versions of the algorithm, demonstrate that intensification aids the performance of EO.

UR - http://www.scopus.com/inward/record.url?scp=78650742187&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-17298-4_12

DO - 10.1007/978-3-642-17298-4_12

M3 - Conference contribution

SN - 3642172970

SN - 9783642172977

VL - 6457 LNCS

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 115

EP - 124

BT - Simulated Evolution and Learning - 8th International Conference, SEAL 2010, Proceedings

ER -

Randall M, Lewis A. Intensification strategies for extremal optimisation. In Simulated Evolution and Learning - 8th International Conference, SEAL 2010, Proceedings. Vol. 6457 LNCS. 2010. p. 115-124. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-17298-4_12