No Cover Image

Conference Paper/Proceeding/Abstract 431 views 58 downloads

Using structural bias to analyse the behaviour of modular CMA-ES

Diederick Vermetten, Fabio Caraffini Orcid Logo, Bas van Stein, Anna V. Kononova

Proceedings of the Genetic and Evolutionary Computation Conference Companion

Swansea University Author: Fabio Caraffini Orcid Logo

  • 62444_VoR.pdf

    PDF | Version of Record

    © 2022 Copyright held by the owner/author(s). Released under the terms of a CC-BY License

    Download (2.36MB)

DOI (Published version): 10.1145/3520304.3534035

Abstract

The Covariance Matrix Adaptation Evolution Strategy (CMA-ES) is a commonly used iterative optimisation heuristic for optimising black-box functions. CMA-ES comes in many flavours with different configuration settings. In this work, we investigate whether CMA-ES suffers from structural bias and which...

Full description

Published in: Proceedings of the Genetic and Evolutionary Computation Conference Companion
ISBN: 978-1-4503-9268-6
Published: New York, NY, USA ACM 2022
URI: https://cronfa.swan.ac.uk/Record/cronfa62444
Tags: Add Tag
No Tags, Be the first to tag this record!
first_indexed 2023-01-25T17:29:00Z
last_indexed 2023-02-21T04:19:04Z
id cronfa62444
recordtype SURis
fullrecord <?xml version="1.0"?><rfc1807><datestamp>2023-02-20T13:33:44.8131335</datestamp><bib-version>v2</bib-version><id>62444</id><entry>2023-01-25</entry><title>Using structural bias to analyse the behaviour of modular CMA-ES</title><swanseaauthors><author><sid>d0b8d4e63d512d4d67a02a23dd20dfdb</sid><ORCID>0000-0001-9199-7368</ORCID><firstname>Fabio</firstname><surname>Caraffini</surname><name>Fabio Caraffini</name><active>true</active><ethesisStudent>false</ethesisStudent></author></swanseaauthors><date>2023-01-25</date><deptcode>SCS</deptcode><abstract>The Covariance Matrix Adaptation Evolution Strategy (CMA-ES) is a commonly used iterative optimisation heuristic for optimising black-box functions. CMA-ES comes in many flavours with different configuration settings. In this work, we investigate whether CMA-ES suffers from structural bias and which modules and parameters affect the strength and type of structural bias. Structural bias occurs when an algorithm or a component of the algorithm biases the search towards a specific direction in the search space irrespective of the objective function. In addition to this investigation, we propose a method to assess the relationship between structural bias and the performance of configurations with different types of bias on the BBOB suite of benchmark functions. Surprisingly for such a popular algorithm, 90.3% of the 1 620 CMA-ES configurations were found to have Structural Bias. Some interesting patterns between module settings and bias types are presented and further insights are discussed.</abstract><type>Conference Paper/Proceeding/Abstract</type><journal>Proceedings of the Genetic and Evolutionary Computation Conference Companion</journal><volume/><journalNumber/><paginationStart/><paginationEnd/><publisher>ACM</publisher><placeOfPublication>New York, NY, USA</placeOfPublication><isbnPrint/><isbnElectronic>978-1-4503-9268-6</isbnElectronic><issnPrint/><issnElectronic/><keywords/><publishedDay>19</publishedDay><publishedMonth>7</publishedMonth><publishedYear>2022</publishedYear><publishedDate>2022-07-19</publishedDate><doi>10.1145/3520304.3534035</doi><url/><notes/><college>COLLEGE NANME</college><department>Computer Science</department><CollegeCode>COLLEGE CODE</CollegeCode><DepartmentCode>SCS</DepartmentCode><institution>Swansea University</institution><apcterm>Another institution paid the OA fee</apcterm><funders/><projectreference/><lastEdited>2023-02-20T13:33:44.8131335</lastEdited><Created>2023-01-25T17:27:17.2232013</Created><path><level id="1">Faculty of Science and Engineering</level><level id="2">School of Mathematics and Computer Science - Computer Science</level></path><authors><author><firstname>Diederick</firstname><surname>Vermetten</surname><order>1</order></author><author><firstname>Fabio</firstname><surname>Caraffini</surname><orcid>0000-0001-9199-7368</orcid><order>2</order></author><author><firstname>Bas van</firstname><surname>Stein</surname><order>3</order></author><author><firstname>Anna V.</firstname><surname>Kononova</surname><order>4</order></author></authors><documents><document><filename>62444__26627__cb843fa9e6824082ac549f3f1353772a.pdf</filename><originalFilename>62444_VoR.pdf</originalFilename><uploaded>2023-02-20T13:24:46.0473030</uploaded><type>Output</type><contentLength>2472372</contentLength><contentType>application/pdf</contentType><version>Version of Record</version><cronfaStatus>true</cronfaStatus><documentNotes>&#xA9; 2022 Copyright held by the owner/author(s). Released under the terms of a CC-BY License</documentNotes><copyrightCorrect>true</copyrightCorrect><language>eng</language><licence>https://creativecommons.org/licenses/by/4.0/</licence></document></documents><OutputDurs/></rfc1807>
spelling 2023-02-20T13:33:44.8131335 v2 62444 2023-01-25 Using structural bias to analyse the behaviour of modular CMA-ES d0b8d4e63d512d4d67a02a23dd20dfdb 0000-0001-9199-7368 Fabio Caraffini Fabio Caraffini true false 2023-01-25 SCS The Covariance Matrix Adaptation Evolution Strategy (CMA-ES) is a commonly used iterative optimisation heuristic for optimising black-box functions. CMA-ES comes in many flavours with different configuration settings. In this work, we investigate whether CMA-ES suffers from structural bias and which modules and parameters affect the strength and type of structural bias. Structural bias occurs when an algorithm or a component of the algorithm biases the search towards a specific direction in the search space irrespective of the objective function. In addition to this investigation, we propose a method to assess the relationship between structural bias and the performance of configurations with different types of bias on the BBOB suite of benchmark functions. Surprisingly for such a popular algorithm, 90.3% of the 1 620 CMA-ES configurations were found to have Structural Bias. Some interesting patterns between module settings and bias types are presented and further insights are discussed. Conference Paper/Proceeding/Abstract Proceedings of the Genetic and Evolutionary Computation Conference Companion ACM New York, NY, USA 978-1-4503-9268-6 19 7 2022 2022-07-19 10.1145/3520304.3534035 COLLEGE NANME Computer Science COLLEGE CODE SCS Swansea University Another institution paid the OA fee 2023-02-20T13:33:44.8131335 2023-01-25T17:27:17.2232013 Faculty of Science and Engineering School of Mathematics and Computer Science - Computer Science Diederick Vermetten 1 Fabio Caraffini 0000-0001-9199-7368 2 Bas van Stein 3 Anna V. Kononova 4 62444__26627__cb843fa9e6824082ac549f3f1353772a.pdf 62444_VoR.pdf 2023-02-20T13:24:46.0473030 Output 2472372 application/pdf Version of Record true © 2022 Copyright held by the owner/author(s). Released under the terms of a CC-BY License true eng https://creativecommons.org/licenses/by/4.0/
title Using structural bias to analyse the behaviour of modular CMA-ES
spellingShingle Using structural bias to analyse the behaviour of modular CMA-ES
Fabio Caraffini
title_short Using structural bias to analyse the behaviour of modular CMA-ES
title_full Using structural bias to analyse the behaviour of modular CMA-ES
title_fullStr Using structural bias to analyse the behaviour of modular CMA-ES
title_full_unstemmed Using structural bias to analyse the behaviour of modular CMA-ES
title_sort Using structural bias to analyse the behaviour of modular CMA-ES
author_id_str_mv d0b8d4e63d512d4d67a02a23dd20dfdb
author_id_fullname_str_mv d0b8d4e63d512d4d67a02a23dd20dfdb_***_Fabio Caraffini
author Fabio Caraffini
author2 Diederick Vermetten
Fabio Caraffini
Bas van Stein
Anna V. Kononova
format Conference Paper/Proceeding/Abstract
container_title Proceedings of the Genetic and Evolutionary Computation Conference Companion
publishDate 2022
institution Swansea University
isbn 978-1-4503-9268-6
doi_str_mv 10.1145/3520304.3534035
publisher ACM
college_str Faculty of Science and Engineering
hierarchytype
hierarchy_top_id facultyofscienceandengineering
hierarchy_top_title Faculty of Science and Engineering
hierarchy_parent_id facultyofscienceandengineering
hierarchy_parent_title Faculty of Science and Engineering
department_str School of Mathematics and Computer Science - Computer Science{{{_:::_}}}Faculty of Science and Engineering{{{_:::_}}}School of Mathematics and Computer Science - Computer Science
document_store_str 1
active_str 0
description The Covariance Matrix Adaptation Evolution Strategy (CMA-ES) is a commonly used iterative optimisation heuristic for optimising black-box functions. CMA-ES comes in many flavours with different configuration settings. In this work, we investigate whether CMA-ES suffers from structural bias and which modules and parameters affect the strength and type of structural bias. Structural bias occurs when an algorithm or a component of the algorithm biases the search towards a specific direction in the search space irrespective of the objective function. In addition to this investigation, we propose a method to assess the relationship between structural bias and the performance of configurations with different types of bias on the BBOB suite of benchmark functions. Surprisingly for such a popular algorithm, 90.3% of the 1 620 CMA-ES configurations were found to have Structural Bias. Some interesting patterns between module settings and bias types are presented and further insights are discussed.
published_date 2022-07-19T04:22:04Z
_version_ 1763754455117856768
score 11.021648

Similar Items