A Surrogate-Assisted Cooperative Co-evolutionary Algorithm Using Recursive Differential Grouping as Decomposition Strategy

Research output: Contribution to journalArticle

Abstract

ooperative co-evolutionary algorithms, especiallythose able to uncover interaction structure between variables,have a great potential in optimizing large-scale problems. Nev-ertheless, they are expensive in terms of number of functionevaluations and this issue can be quite problematic whendealing with computationally expensive optimization problems.An effective approach to deal with such problems lies in theexploitation of surrogate models. The latter ones work as cheap-to-evaluate alternatives to the expensive function reducing thecomputational cost, while still providing improved designs. Thisprocess, called surrogate-assisted optimization, is very effective onsmall-dimensional problems but is not suitable to solve large-scaleproblems due to the curse of dimensionality. In this paper, a newalgorithm, taking benefit from cooperative coevolution and sur-rogate models, is introduced to efficiently solve high-dimensional,expensive and black-box problems. The proposed algorithm usesrecursive differential grouping to perform an accurate problemdecomposition. Experimental results are provided on a set of1000-dimensional problems and show promising results
Original languageEnglish
Article number18888677
Pages (from-to)674-681
Number of pages8
Journal2019 IEEE Congress on Evolutionary Computation (CEC)
DOIs
Publication statusPublished - 8 Aug 2019
Event2019 IEEE Congress On Evolutionary Computation - Te Papa Tongarewa, Wellington, New Zealand
Duration: 10 Jun 201913 Jun 2019
http://cec2019.org/

Fingerprint

Evolutionary algorithms
Decomposition
Costs

Keywords

  • global optimization
  • surrogate-assisted optimization
  • large-scale optimization
  • high dimensional
  • expensive and black-box problems
  • cooperative co-evolutionary algorithm
  • differential grouping
  • evolutionary algorithm

Cite this

@article{14d302281b9941369517bb43cf15ae3d,
title = "A Surrogate-Assisted Cooperative Co-evolutionary Algorithm Using Recursive Differential Grouping as Decomposition Strategy",
abstract = "ooperative co-evolutionary algorithms, especiallythose able to uncover interaction structure between variables,have a great potential in optimizing large-scale problems. Nev-ertheless, they are expensive in terms of number of functionevaluations and this issue can be quite problematic whendealing with computationally expensive optimization problems.An effective approach to deal with such problems lies in theexploitation of surrogate models. The latter ones work as cheap-to-evaluate alternatives to the expensive function reducing thecomputational cost, while still providing improved designs. Thisprocess, called surrogate-assisted optimization, is very effective onsmall-dimensional problems but is not suitable to solve large-scaleproblems due to the curse of dimensionality. In this paper, a newalgorithm, taking benefit from cooperative coevolution and sur-rogate models, is introduced to efficiently solve high-dimensional,expensive and black-box problems. The proposed algorithm usesrecursive differential grouping to perform an accurate problemdecomposition. Experimental results are provided on a set of1000-dimensional problems and show promising results",
keywords = "global optimization, surrogate-assisted optimization, Large-scale optimization, high dimensional , expensive and black-box problems, cooperative co-evolutionary algorithm, differential grouping, evolutionary algorithm, global optimization, surrogate-assisted optimization, large-scale optimization, high dimensional, expensive and black-box problems, cooperative co-evolutionary algorithm, differential grouping, evolutionary algorithm",
author = "Julien Blanchard and Charlotte Beauthier and Timoteo Carletti",
year = "2019",
month = "8",
day = "8",
doi = "10.1109/CEC.2019.8790114",
language = "English",
pages = "674--681",
journal = "2019 IEEE Congress on Evolutionary Computation (CEC)",
publisher = "IEEE",

}

TY - JOUR

T1 - A Surrogate-Assisted Cooperative Co-evolutionary Algorithm Using Recursive Differential Grouping as Decomposition Strategy

AU - Blanchard, Julien

AU - Beauthier, Charlotte

AU - Carletti, Timoteo

PY - 2019/8/8

Y1 - 2019/8/8

N2 - ooperative co-evolutionary algorithms, especiallythose able to uncover interaction structure between variables,have a great potential in optimizing large-scale problems. Nev-ertheless, they are expensive in terms of number of functionevaluations and this issue can be quite problematic whendealing with computationally expensive optimization problems.An effective approach to deal with such problems lies in theexploitation of surrogate models. The latter ones work as cheap-to-evaluate alternatives to the expensive function reducing thecomputational cost, while still providing improved designs. Thisprocess, called surrogate-assisted optimization, is very effective onsmall-dimensional problems but is not suitable to solve large-scaleproblems due to the curse of dimensionality. In this paper, a newalgorithm, taking benefit from cooperative coevolution and sur-rogate models, is introduced to efficiently solve high-dimensional,expensive and black-box problems. The proposed algorithm usesrecursive differential grouping to perform an accurate problemdecomposition. Experimental results are provided on a set of1000-dimensional problems and show promising results

AB - ooperative co-evolutionary algorithms, especiallythose able to uncover interaction structure between variables,have a great potential in optimizing large-scale problems. Nev-ertheless, they are expensive in terms of number of functionevaluations and this issue can be quite problematic whendealing with computationally expensive optimization problems.An effective approach to deal with such problems lies in theexploitation of surrogate models. The latter ones work as cheap-to-evaluate alternatives to the expensive function reducing thecomputational cost, while still providing improved designs. Thisprocess, called surrogate-assisted optimization, is very effective onsmall-dimensional problems but is not suitable to solve large-scaleproblems due to the curse of dimensionality. In this paper, a newalgorithm, taking benefit from cooperative coevolution and sur-rogate models, is introduced to efficiently solve high-dimensional,expensive and black-box problems. The proposed algorithm usesrecursive differential grouping to perform an accurate problemdecomposition. Experimental results are provided on a set of1000-dimensional problems and show promising results

KW - global optimization

KW - surrogate-assisted optimization

KW - Large-scale optimization

KW - high dimensional , expensive and black-box problems

KW - cooperative co-evolutionary algorithm

KW - differential grouping

KW - evolutionary algorithm

KW - global optimization

KW - surrogate-assisted optimization

KW - large-scale optimization

KW - high dimensional

KW - expensive and black-box problems

KW - cooperative co-evolutionary algorithm

KW - differential grouping

KW - evolutionary algorithm

UR - http://www.scopus.com/inward/record.url?scp=85071327186&partnerID=8YFLogxK

U2 - 10.1109/CEC.2019.8790114

DO - 10.1109/CEC.2019.8790114

M3 - Article

SP - 674

EP - 681

JO - 2019 IEEE Congress on Evolutionary Computation (CEC)

JF - 2019 IEEE Congress on Evolutionary Computation (CEC)

M1 - 18888677

ER -