No Cover Image

Conference Paper/Proceeding/Abstract 716 views 151 downloads

ϵ-shotgun: ϵ-greedy batch bayesian optimisation

George De Ath, Richard M. Everson, Jonathan E. Fieldsend, Alma Rahat Orcid Logo

Proceedings of the 2020 Genetic and Evolutionary Computation Conference, Pages: 787 - 795

Swansea University Author: Alma Rahat Orcid Logo

DOI (Published version): 10.1145/3377930.3390154

Abstract

Bayesian optimisation is a popular surrogate model-based approach for optimising expensive black-box functions. Given a surrogate model, the next location to expensively evaluate is chosen via maximisation of a cheap-to-query acquisition function. We present an ϵ-greedy procedure for Bayesian optimi...

Full description

Published in: Proceedings of the 2020 Genetic and Evolutionary Computation Conference
ISBN: 9781450371285
Published: New York, NY, USA ACM 2020
URI: https://cronfa.swan.ac.uk/Record/cronfa54660
Tags: Add Tag
No Tags, Be the first to tag this record!
first_indexed 2020-07-07T07:38:23Z
last_indexed 2020-11-28T04:10:31Z
id cronfa54660
recordtype SURis
fullrecord <?xml version="1.0"?><rfc1807><datestamp>2020-11-27T11:50:42.8033895</datestamp><bib-version>v2</bib-version><id>54660</id><entry>2020-07-07</entry><title>&#x3F5;-shotgun: &#x3F5;-greedy batch bayesian optimisation</title><swanseaauthors><author><sid>6206f027aca1e3a5ff6b8cd224248bc2</sid><ORCID>0000-0002-5023-1371</ORCID><firstname>Alma</firstname><surname>Rahat</surname><name>Alma Rahat</name><active>true</active><ethesisStudent>false</ethesisStudent></author></swanseaauthors><date>2020-07-07</date><deptcode>SCS</deptcode><abstract>Bayesian optimisation is a popular surrogate model-based approach for optimising expensive black-box functions. Given a surrogate model, the next location to expensively evaluate is chosen via maximisation of a cheap-to-query acquisition function. We present an &#x3F5;-greedy procedure for Bayesian optimisation in batch settings in which the black-box function can be evaluated multiple times in parallel. Our &#x3F5;-shotgun algorithm leverages the model's prediction, uncertainty, and the approximated rate of change of the landscape to determine the spread of batch solutions to be distributed around a putative location. The initial target location is selected either in an exploitative fashion on the mean prediction, or - with probability &#x3F5; - from elsewhere in the design space. This results in locations that are more densely sampled in regions where the function is changing rapidly and in locations predicted to be good (i.e. close to predicted optima), with more scattered samples in regions where the function is flatter and/or of poorer quality. We empirically evaluate the &#x3F5;-shotgun methods on a range of synthetic functions and two real-world problems, finding that they perform at least as well as state-of-the-art batch methods and in many cases exceed their performance.</abstract><type>Conference Paper/Proceeding/Abstract</type><journal>Proceedings of the 2020 Genetic and Evolutionary Computation Conference</journal><volume/><journalNumber/><paginationStart>787</paginationStart><paginationEnd>795</paginationEnd><publisher>ACM</publisher><placeOfPublication>New York, NY, USA</placeOfPublication><isbnPrint>9781450371285</isbnPrint><isbnElectronic/><issnPrint/><issnElectronic/><keywords>Bayesian optimisation, Batch, Parallel, Exploitation, -greedy, Infill criteria, Acquisition function</keywords><publishedDay>26</publishedDay><publishedMonth>6</publishedMonth><publishedYear>2020</publishedYear><publishedDate>2020-06-26</publishedDate><doi>10.1145/3377930.3390154</doi><url/><notes/><college>COLLEGE NANME</college><department>Computer Science</department><CollegeCode>COLLEGE CODE</CollegeCode><DepartmentCode>SCS</DepartmentCode><institution>Swansea University</institution><apcterm/><lastEdited>2020-11-27T11:50:42.8033895</lastEdited><Created>2020-07-07T08:28:12.7698158</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>George De</firstname><surname>Ath</surname><order>1</order></author><author><firstname>Richard M.</firstname><surname>Everson</surname><order>2</order></author><author><firstname>Jonathan E.</firstname><surname>Fieldsend</surname><order>3</order></author><author><firstname>Alma</firstname><surname>Rahat</surname><orcid>0000-0002-5023-1371</orcid><order>4</order></author></authors><documents><document><filename>54660__17657__bfc7bcce6f5b47eb9ebac4bb4d03a248.pdf</filename><originalFilename>BBO_peaks.pdf</originalFilename><uploaded>2020-07-07T08:37:57.2810048</uploaded><type>Output</type><contentLength>3689462</contentLength><contentType>application/pdf</contentType><version>Accepted Manuscript</version><cronfaStatus>true</cronfaStatus><copyrightCorrect>true</copyrightCorrect><language>eng</language></document></documents><OutputDurs/></rfc1807>
spelling 2020-11-27T11:50:42.8033895 v2 54660 2020-07-07 ϵ-shotgun: ϵ-greedy batch bayesian optimisation 6206f027aca1e3a5ff6b8cd224248bc2 0000-0002-5023-1371 Alma Rahat Alma Rahat true false 2020-07-07 SCS Bayesian optimisation is a popular surrogate model-based approach for optimising expensive black-box functions. Given a surrogate model, the next location to expensively evaluate is chosen via maximisation of a cheap-to-query acquisition function. We present an ϵ-greedy procedure for Bayesian optimisation in batch settings in which the black-box function can be evaluated multiple times in parallel. Our ϵ-shotgun algorithm leverages the model's prediction, uncertainty, and the approximated rate of change of the landscape to determine the spread of batch solutions to be distributed around a putative location. The initial target location is selected either in an exploitative fashion on the mean prediction, or - with probability ϵ - from elsewhere in the design space. This results in locations that are more densely sampled in regions where the function is changing rapidly and in locations predicted to be good (i.e. close to predicted optima), with more scattered samples in regions where the function is flatter and/or of poorer quality. We empirically evaluate the ϵ-shotgun methods on a range of synthetic functions and two real-world problems, finding that they perform at least as well as state-of-the-art batch methods and in many cases exceed their performance. Conference Paper/Proceeding/Abstract Proceedings of the 2020 Genetic and Evolutionary Computation Conference 787 795 ACM New York, NY, USA 9781450371285 Bayesian optimisation, Batch, Parallel, Exploitation, -greedy, Infill criteria, Acquisition function 26 6 2020 2020-06-26 10.1145/3377930.3390154 COLLEGE NANME Computer Science COLLEGE CODE SCS Swansea University 2020-11-27T11:50:42.8033895 2020-07-07T08:28:12.7698158 Faculty of Science and Engineering School of Mathematics and Computer Science - Computer Science George De Ath 1 Richard M. Everson 2 Jonathan E. Fieldsend 3 Alma Rahat 0000-0002-5023-1371 4 54660__17657__bfc7bcce6f5b47eb9ebac4bb4d03a248.pdf BBO_peaks.pdf 2020-07-07T08:37:57.2810048 Output 3689462 application/pdf Accepted Manuscript true true eng
title ϵ-shotgun: ϵ-greedy batch bayesian optimisation
spellingShingle ϵ-shotgun: ϵ-greedy batch bayesian optimisation
Alma Rahat
title_short ϵ-shotgun: ϵ-greedy batch bayesian optimisation
title_full ϵ-shotgun: ϵ-greedy batch bayesian optimisation
title_fullStr ϵ-shotgun: ϵ-greedy batch bayesian optimisation
title_full_unstemmed ϵ-shotgun: ϵ-greedy batch bayesian optimisation
title_sort ϵ-shotgun: ϵ-greedy batch bayesian optimisation
author_id_str_mv 6206f027aca1e3a5ff6b8cd224248bc2
author_id_fullname_str_mv 6206f027aca1e3a5ff6b8cd224248bc2_***_Alma Rahat
author Alma Rahat
author2 George De Ath
Richard M. Everson
Jonathan E. Fieldsend
Alma Rahat
format Conference Paper/Proceeding/Abstract
container_title Proceedings of the 2020 Genetic and Evolutionary Computation Conference
container_start_page 787
publishDate 2020
institution Swansea University
isbn 9781450371285
doi_str_mv 10.1145/3377930.3390154
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 Bayesian optimisation is a popular surrogate model-based approach for optimising expensive black-box functions. Given a surrogate model, the next location to expensively evaluate is chosen via maximisation of a cheap-to-query acquisition function. We present an ϵ-greedy procedure for Bayesian optimisation in batch settings in which the black-box function can be evaluated multiple times in parallel. Our ϵ-shotgun algorithm leverages the model's prediction, uncertainty, and the approximated rate of change of the landscape to determine the spread of batch solutions to be distributed around a putative location. The initial target location is selected either in an exploitative fashion on the mean prediction, or - with probability ϵ - from elsewhere in the design space. This results in locations that are more densely sampled in regions where the function is changing rapidly and in locations predicted to be good (i.e. close to predicted optima), with more scattered samples in regions where the function is flatter and/or of poorer quality. We empirically evaluate the ϵ-shotgun methods on a range of synthetic functions and two real-world problems, finding that they perform at least as well as state-of-the-art batch methods and in many cases exceed their performance.
published_date 2020-06-26T04:08:19Z
_version_ 1763753590714793984
score 11.012678

Similar Items