Analysis of emergent selection pressure in evolutionary algorithm and machine learner offspring filtering hybrids

Mark Coletti, Guido Cervone

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

Abstract

When evolutionary algorithms are applied to problems with computationally intensive fitness functions a limited budget of evaluations is usually available. For these types of problems minimizing the number of function evaluations becomes paramount, which can be achieved by using smaller population sizes and limiting the number of generations per run. Unfortunately this leads to a limited sampling of the problem space, which means finding adequate solutions is less likely. Evolutionary algorithms (EA) can be augmented with machine learners (ML) to more effectively explore the problem space. However, a "well-tuned" evolutionary algorithm strikes a balance between its constituent operators. Failure to do so could mean implementations that prematurely converge to inferior solutions or to not converge at all. One aspect of such "tuning" is the use of a proper selection pressure. Introducing a machine learner into an EA/ML hybrid introduces a new form of "emergent" selection pressure for which practitioners may need to compensate. This research shows two implementations of EA/ML hybrids that filter out inferior offspring based on knowledge inferred from better individuals have different emergent selection pressure characteristics.

Original languageEnglish (US)
Title of host publicationSwarm, Evolutionary, and Memetic Computing - Third International Conference, SEMCCO 2012, Proceedings
Pages721-728
Number of pages8
DOIs
StatePublished - Dec 31 2012
Event3rd International Conference on Swarm, Evolutionary, and Memetic Computing, SEMCCO 2012 - Bhubaneswar, India
Duration: Dec 20 2012Dec 22 2012

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume7677 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other3rd International Conference on Swarm, Evolutionary, and Memetic Computing, SEMCCO 2012
CountryIndia
CityBhubaneswar
Period12/20/1212/22/12

Fingerprint

Evolutionary algorithms
Evolutionary Algorithms
Filtering
Converge
Function evaluation
Evaluation Function
Fitness Function
Population Size
Mathematical operators
Tuning
Limiting
Likely
Filter
Sampling
Evaluation
Operator

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Coletti, M., & Cervone, G. (2012). Analysis of emergent selection pressure in evolutionary algorithm and machine learner offspring filtering hybrids. In Swarm, Evolutionary, and Memetic Computing - Third International Conference, SEMCCO 2012, Proceedings (pp. 721-728). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7677 LNCS). https://doi.org/10.1007/978-3-642-35380-2_84
Coletti, Mark ; Cervone, Guido. / Analysis of emergent selection pressure in evolutionary algorithm and machine learner offspring filtering hybrids. Swarm, Evolutionary, and Memetic Computing - Third International Conference, SEMCCO 2012, Proceedings. 2012. pp. 721-728 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{191ccc0bfe7e468493e8856a837c06b6,
title = "Analysis of emergent selection pressure in evolutionary algorithm and machine learner offspring filtering hybrids",
abstract = "When evolutionary algorithms are applied to problems with computationally intensive fitness functions a limited budget of evaluations is usually available. For these types of problems minimizing the number of function evaluations becomes paramount, which can be achieved by using smaller population sizes and limiting the number of generations per run. Unfortunately this leads to a limited sampling of the problem space, which means finding adequate solutions is less likely. Evolutionary algorithms (EA) can be augmented with machine learners (ML) to more effectively explore the problem space. However, a {"}well-tuned{"} evolutionary algorithm strikes a balance between its constituent operators. Failure to do so could mean implementations that prematurely converge to inferior solutions or to not converge at all. One aspect of such {"}tuning{"} is the use of a proper selection pressure. Introducing a machine learner into an EA/ML hybrid introduces a new form of {"}emergent{"} selection pressure for which practitioners may need to compensate. This research shows two implementations of EA/ML hybrids that filter out inferior offspring based on knowledge inferred from better individuals have different emergent selection pressure characteristics.",
author = "Mark Coletti and Guido Cervone",
year = "2012",
month = "12",
day = "31",
doi = "10.1007/978-3-642-35380-2_84",
language = "English (US)",
isbn = "9783642353796",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "721--728",
booktitle = "Swarm, Evolutionary, and Memetic Computing - Third International Conference, SEMCCO 2012, Proceedings",

}

Coletti, M & Cervone, G 2012, Analysis of emergent selection pressure in evolutionary algorithm and machine learner offspring filtering hybrids. in Swarm, Evolutionary, and Memetic Computing - Third International Conference, SEMCCO 2012, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 7677 LNCS, pp. 721-728, 3rd International Conference on Swarm, Evolutionary, and Memetic Computing, SEMCCO 2012, Bhubaneswar, India, 12/20/12. https://doi.org/10.1007/978-3-642-35380-2_84

Analysis of emergent selection pressure in evolutionary algorithm and machine learner offspring filtering hybrids. / Coletti, Mark; Cervone, Guido.

Swarm, Evolutionary, and Memetic Computing - Third International Conference, SEMCCO 2012, Proceedings. 2012. p. 721-728 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7677 LNCS).

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

TY - GEN

T1 - Analysis of emergent selection pressure in evolutionary algorithm and machine learner offspring filtering hybrids

AU - Coletti, Mark

AU - Cervone, Guido

PY - 2012/12/31

Y1 - 2012/12/31

N2 - When evolutionary algorithms are applied to problems with computationally intensive fitness functions a limited budget of evaluations is usually available. For these types of problems minimizing the number of function evaluations becomes paramount, which can be achieved by using smaller population sizes and limiting the number of generations per run. Unfortunately this leads to a limited sampling of the problem space, which means finding adequate solutions is less likely. Evolutionary algorithms (EA) can be augmented with machine learners (ML) to more effectively explore the problem space. However, a "well-tuned" evolutionary algorithm strikes a balance between its constituent operators. Failure to do so could mean implementations that prematurely converge to inferior solutions or to not converge at all. One aspect of such "tuning" is the use of a proper selection pressure. Introducing a machine learner into an EA/ML hybrid introduces a new form of "emergent" selection pressure for which practitioners may need to compensate. This research shows two implementations of EA/ML hybrids that filter out inferior offspring based on knowledge inferred from better individuals have different emergent selection pressure characteristics.

AB - When evolutionary algorithms are applied to problems with computationally intensive fitness functions a limited budget of evaluations is usually available. For these types of problems minimizing the number of function evaluations becomes paramount, which can be achieved by using smaller population sizes and limiting the number of generations per run. Unfortunately this leads to a limited sampling of the problem space, which means finding adequate solutions is less likely. Evolutionary algorithms (EA) can be augmented with machine learners (ML) to more effectively explore the problem space. However, a "well-tuned" evolutionary algorithm strikes a balance between its constituent operators. Failure to do so could mean implementations that prematurely converge to inferior solutions or to not converge at all. One aspect of such "tuning" is the use of a proper selection pressure. Introducing a machine learner into an EA/ML hybrid introduces a new form of "emergent" selection pressure for which practitioners may need to compensate. This research shows two implementations of EA/ML hybrids that filter out inferior offspring based on knowledge inferred from better individuals have different emergent selection pressure characteristics.

UR - http://www.scopus.com/inward/record.url?scp=84871569153&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84871569153&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-35380-2_84

DO - 10.1007/978-3-642-35380-2_84

M3 - Conference contribution

SN - 9783642353796

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

SP - 721

EP - 728

BT - Swarm, Evolutionary, and Memetic Computing - Third International Conference, SEMCCO 2012, Proceedings

ER -

Coletti M, Cervone G. Analysis of emergent selection pressure in evolutionary algorithm and machine learner offspring filtering hybrids. In Swarm, Evolutionary, and Memetic Computing - Third International Conference, SEMCCO 2012, Proceedings. 2012. p. 721-728. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-35380-2_84