Adaptive pattern search for large-scale optimization

Vincent Gardeux, Mahamed G. Mahamed, Rachid Chelouah, Patrick Siarry, Fred Glover

Research output: Research - peer-reviewArticle

Abstract

The emergence of high-dimensional data requires the design of new optimization methods. Indeed, conventional optimization methods require improvements, hybridization, or parameter tuning in order to operate in spaces of high dimensions. In this paper, we present a new adaptive variant of a pattern search algorithm to solve global optimization problems exhibiting such a character. The proposed method has no parameters visible to the user and the default settings, determined by almost no a priori experimentation, are highly robust on the tested datasets. The algorithm is evaluated and compared with 11 state-of-the-art methods on 20 benchmark functions of 1000 dimensions from the CEC’2010 competition. The results show that this approach obtains good performances compared to the other methods tested.

LanguageEnglish (US)
Pages1-12
Number of pages12
JournalApplied Intelligence
DOIs
StateAccepted/In press - Mar 11 2017
Externally publishedYes

Fingerprint

Global optimization
Tuning

Keywords

  • Adaptive methods
  • Continuous
  • High-dimension
  • Large-scale
  • Optimization
  • Pattern search
  • Scatter search

ASJC Scopus subject areas

  • Artificial Intelligence

Cite this

Gardeux, V., Mahamed, M. G., Chelouah, R., Siarry, P., & Glover, F. (2017). Adaptive pattern search for large-scale optimization. Applied Intelligence, 1-12. DOI: 10.1007/s10489-017-0901-8

Adaptive pattern search for large-scale optimization. / Gardeux, Vincent; Mahamed, Mahamed G.; Chelouah, Rachid; Siarry, Patrick; Glover, Fred.

In: Applied Intelligence, 11.03.2017, p. 1-12.

Research output: Research - peer-reviewArticle

Gardeux, V, Mahamed, MG, Chelouah, R, Siarry, P & Glover, F 2017, 'Adaptive pattern search for large-scale optimization' Applied Intelligence, pp. 1-12. DOI: 10.1007/s10489-017-0901-8
Gardeux V, Mahamed MG, Chelouah R, Siarry P, Glover F. Adaptive pattern search for large-scale optimization. Applied Intelligence. 2017 Mar 11;1-12. Available from, DOI: 10.1007/s10489-017-0901-8
Gardeux, Vincent ; Mahamed, Mahamed G. ; Chelouah, Rachid ; Siarry, Patrick ; Glover, Fred. / Adaptive pattern search for large-scale optimization. In: Applied Intelligence. 2017 ; pp. 1-12
@article{60f29dd3166c4d8ebf84469b8040ba7d,
title = "Adaptive pattern search for large-scale optimization",
abstract = "The emergence of high-dimensional data requires the design of new optimization methods. Indeed, conventional optimization methods require improvements, hybridization, or parameter tuning in order to operate in spaces of high dimensions. In this paper, we present a new adaptive variant of a pattern search algorithm to solve global optimization problems exhibiting such a character. The proposed method has no parameters visible to the user and the default settings, determined by almost no a priori experimentation, are highly robust on the tested datasets. The algorithm is evaluated and compared with 11 state-of-the-art methods on 20 benchmark functions of 1000 dimensions from the CEC’2010 competition. The results show that this approach obtains good performances compared to the other methods tested.",
keywords = "Adaptive methods, Continuous, High-dimension, Large-scale, Optimization, Pattern search, Scatter search",
author = "Vincent Gardeux and Mahamed, {Mahamed G.} and Rachid Chelouah and Patrick Siarry and Fred Glover",
year = "2017",
month = "3",
doi = "10.1007/s10489-017-0901-8",
pages = "1--12",
journal = "Applied Intelligence",
issn = "0924-669X",
publisher = "Springer Netherlands",

}

TY - JOUR

T1 - Adaptive pattern search for large-scale optimization

AU - Gardeux,Vincent

AU - Mahamed,Mahamed G.

AU - Chelouah,Rachid

AU - Siarry,Patrick

AU - Glover,Fred

PY - 2017/3/11

Y1 - 2017/3/11

N2 - The emergence of high-dimensional data requires the design of new optimization methods. Indeed, conventional optimization methods require improvements, hybridization, or parameter tuning in order to operate in spaces of high dimensions. In this paper, we present a new adaptive variant of a pattern search algorithm to solve global optimization problems exhibiting such a character. The proposed method has no parameters visible to the user and the default settings, determined by almost no a priori experimentation, are highly robust on the tested datasets. The algorithm is evaluated and compared with 11 state-of-the-art methods on 20 benchmark functions of 1000 dimensions from the CEC’2010 competition. The results show that this approach obtains good performances compared to the other methods tested.

AB - The emergence of high-dimensional data requires the design of new optimization methods. Indeed, conventional optimization methods require improvements, hybridization, or parameter tuning in order to operate in spaces of high dimensions. In this paper, we present a new adaptive variant of a pattern search algorithm to solve global optimization problems exhibiting such a character. The proposed method has no parameters visible to the user and the default settings, determined by almost no a priori experimentation, are highly robust on the tested datasets. The algorithm is evaluated and compared with 11 state-of-the-art methods on 20 benchmark functions of 1000 dimensions from the CEC’2010 competition. The results show that this approach obtains good performances compared to the other methods tested.

KW - Adaptive methods

KW - Continuous

KW - High-dimension

KW - Large-scale

KW - Optimization

KW - Pattern search

KW - Scatter search

UR - http://www.scopus.com/inward/record.url?scp=85014750353&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85014750353&partnerID=8YFLogxK

U2 - 10.1007/s10489-017-0901-8

DO - 10.1007/s10489-017-0901-8

M3 - Article

SP - 1

EP - 12

JO - Applied Intelligence

T2 - Applied Intelligence

JF - Applied Intelligence

SN - 0924-669X

ER -