Differential evolution for strongly noisy optimization: Use 1.01n resamplings at iteration n and Reach the -1/2 slope

Shih Yuan CHIU; Ching Nung LIN; Jialin LIU; Tsan-Cheng SU; Fabien TEYTAUD; Olivier TEYTAUD; Shi Jim YEN

doi:10.1109/CEC.2015.7256911

Differential evolution for strongly noisy optimization: Use 1.01n resamplings at iteration n and Reach the -1/2 slope

Shih Yuan CHIU, Ching Nung LIN, Jialin LIU, Tsan-Cheng SU, Fabien TEYTAUD, Olivier TEYTAUD, Shi Jim YEN

Research output: Book Chapters | Papers in Conference Proceedings › Conference paper (refereed) › Referred Conference Paper › peer-review

6 Citations (Scopus)

Abstract

This paper is devoted to noisy optimization in case of a noise with standard deviation as large as variations of the fitness values, specifically when the variance does not decrease to zero around the optimum. We focus on comparing methods for choosing the number of resamplings. Experiments are performed on the differential evolution algorithm. By mathematical analysis, we design a new rule for choosing the number of resamplings for noisy optimization, as a function of the dimension, and validate its efficiency compared to existing heuristics.

Original language	English
Title of host publication	2015 IEEE Congress on Evolutionary Computation, CEC 2015 : Proceedings
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	338-345
Number of pages	8
ISBN (Electronic)	9781479974924
DOIs	https://doi.org/10.1109/CEC.2015.7256911
Publication status	Published - 2015
Externally published	Yes
Event	IEEE Congress on Evolutionary Computation, CEC 2015 - Sendai, Japan Duration: 25 May 2015 → 28 May 2015

Conference

Conference	IEEE Congress on Evolutionary Computation, CEC 2015
Country/Territory	Japan
City	Sendai
Period	25/05/15 → 28/05/15

Bibliographical note

Publisher Copyright:
© 2015 IEEE.

Keywords

Differential Evolution
Noisy Optimization
Resampling

Access to Document

10.1109/CEC.2015.7256911

Cite this

CHIU, S. Y., LIN, C. N., LIU, J., SU, T.-C., TEYTAUD, F., TEYTAUD, O., & YEN, S. J. (2015). Differential evolution for strongly noisy optimization: Use 1.01n resamplings at iteration n and Reach the -1/2 slope. In 2015 IEEE Congress on Evolutionary Computation, CEC 2015 : Proceedings (pp. 338-345). Article 7256911 Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/CEC.2015.7256911

@inproceedings{2191295890ac4a56a5f863ee52d2194e,

title = "Differential evolution for strongly noisy optimization: Use 1.01n resamplings at iteration n and Reach the -1/2 slope",

abstract = "This paper is devoted to noisy optimization in case of a noise with standard deviation as large as variations of the fitness values, specifically when the variance does not decrease to zero around the optimum. We focus on comparing methods for choosing the number of resamplings. Experiments are performed on the differential evolution algorithm. By mathematical analysis, we design a new rule for choosing the number of resamplings for noisy optimization, as a function of the dimension, and validate its efficiency compared to existing heuristics.",

keywords = "Differential Evolution, Noisy Optimization, Resampling",

author = "CHIU, {Shih Yuan} and LIN, {Ching Nung} and Jialin LIU and Tsan-Cheng SU and Fabien TEYTAUD and Olivier TEYTAUD and YEN, {Shi Jim}",

note = "Publisher Copyright: {\textcopyright} 2015 IEEE.; IEEE Congress on Evolutionary Computation, CEC 2015 ; Conference date: 25-05-2015 Through 28-05-2015",

year = "2015",

doi = "10.1109/CEC.2015.7256911",

language = "English",

pages = "338--345",

booktitle = "2015 IEEE Congress on Evolutionary Computation, CEC 2015 : Proceedings",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

address = "United States",

}

CHIU, SY, LIN, CN, LIU, J, SU, T-C, TEYTAUD, F, TEYTAUD, O & YEN, SJ 2015, Differential evolution for strongly noisy optimization: Use 1.01n resamplings at iteration n and Reach the -1/2 slope. in 2015 IEEE Congress on Evolutionary Computation, CEC 2015 : Proceedings., 7256911, Institute of Electrical and Electronics Engineers Inc., pp. 338-345, IEEE Congress on Evolutionary Computation, CEC 2015, Sendai, Japan, 25/05/15. https://doi.org/10.1109/CEC.2015.7256911

Differential evolution for strongly noisy optimization: Use 1.01n resamplings at iteration n and Reach the -1/2 slope. / CHIU, Shih Yuan; LIN, Ching Nung; LIU, Jialin et al.
2015 IEEE Congress on Evolutionary Computation, CEC 2015 : Proceedings. Institute of Electrical and Electronics Engineers Inc., 2015. p. 338-345 7256911.

Research output: Book Chapters | Papers in Conference Proceedings › Conference paper (refereed) › Referred Conference Paper › peer-review

TY - GEN

T1 - Differential evolution for strongly noisy optimization: Use 1.01n resamplings at iteration n and Reach the -1/2 slope

AU - CHIU, Shih Yuan

AU - LIN, Ching Nung

AU - LIU, Jialin

AU - SU, Tsan-Cheng

AU - TEYTAUD, Fabien

AU - TEYTAUD, Olivier

AU - YEN, Shi Jim

PY - 2015

Y1 - 2015

N2 - This paper is devoted to noisy optimization in case of a noise with standard deviation as large as variations of the fitness values, specifically when the variance does not decrease to zero around the optimum. We focus on comparing methods for choosing the number of resamplings. Experiments are performed on the differential evolution algorithm. By mathematical analysis, we design a new rule for choosing the number of resamplings for noisy optimization, as a function of the dimension, and validate its efficiency compared to existing heuristics.

AB - This paper is devoted to noisy optimization in case of a noise with standard deviation as large as variations of the fitness values, specifically when the variance does not decrease to zero around the optimum. We focus on comparing methods for choosing the number of resamplings. Experiments are performed on the differential evolution algorithm. By mathematical analysis, we design a new rule for choosing the number of resamplings for noisy optimization, as a function of the dimension, and validate its efficiency compared to existing heuristics.

KW - Differential Evolution

KW - Noisy Optimization

KW - Resampling

UR - http://www.scopus.com/inward/record.url?scp=84963549018&partnerID=8YFLogxK

U2 - 10.1109/CEC.2015.7256911

DO - 10.1109/CEC.2015.7256911

M3 - Conference paper (refereed)

AN - SCOPUS:84963549018

SP - 338

EP - 345

BT - 2015 IEEE Congress on Evolutionary Computation, CEC 2015 : Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - IEEE Congress on Evolutionary Computation, CEC 2015

Y2 - 25 May 2015 through 28 May 2015

ER -

CHIU SY, LIN CN, LIU J, SU TC, TEYTAUD F, TEYTAUD O et al. Differential evolution for strongly noisy optimization: Use 1.01n resamplings at iteration n and Reach the -1/2 slope. In 2015 IEEE Congress on Evolutionary Computation, CEC 2015 : Proceedings. Institute of Electrical and Electronics Engineers Inc. 2015. p. 338-345. 7256911 Epub 2015 Sept 14. doi: 10.1109/CEC.2015.7256911

Differential evolution for strongly noisy optimization: Use 1.01n resamplings at iteration n and Reach the -1/2 slope

Abstract

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this