Publicação

INCORPORATING RADIAL BASIS FUNCTIONS IN GLOBAL AND LOCAL DIRECT SEARCH

Ver documento

Detalhes bibliográficos
Resumo:GLODS is a global derivative-free optimization algorithm, relying on local directional direct search, aided by a clever multistart strategy that does not conduct all the lines of search until the end. In 2015, time of the first release of the corresponding solver, GLODS was shown to be competitive when compared to state-of-the-art algorithms, such as MCS or DIRECT. GLODS resorts to sampling techniques to look for minima on a global scale, not taking advantage of the information gathered in previous iterations. As such, the main goal of this work is to replace the pseudo-random sampling approach, used by GLODS to initialize new lines of search, by the minimization of global models of the objective function, defined using radial basis functions, and computed using the points previously evaluated by the algorithm. This should allow a better placement of the starting points for new local lines of search, and, in turn, significantly increase the numerical performance of the algorithm. Naturally, incorporating radial basis functions in GLODS poses new challenges. In this work, we will address questions such as which radial basis functions to use, which points should be selected to compute them, how to minimize these functions, and how to take advantage of their minima in the execution of the algorithm. The new version of GLODS, incorporating radial basis functions, was calibrated to its best numerical performance, and then compared against other state-of-the-art solvers, such as MCS, DIRECT, MATSuMoTo, and ZOOpt. The results obtained are strongly positive. The new algorithm clearly outperforms its previous version, and is competitive with the other solvers tested. Finally, parallel strategies were implemented and tested. Results showed that it is very beneficial to evaluate multiple points simultaneously, for objective functions whose evaluation time is as low as 0.1 seconds. The proposed algorithm, called BoostGLODS, is a cutting-edge, powerful and efficient parallel global derivative-free optimization algo- rithm.
Autores principais:Baptista, Bruno Alexandre da Anunciação
Assunto:Global optimization Derivative-free optimization Radial basis functions Surrogate models Direct search methods Pattern search methods
Ano:2022
País:Portugal
Tipo de documento:dissertação de mestrado
Tipo de acesso:acesso aberto
Instituição associada:Universidade Nova de Lisboa
Idioma:inglês
Origem:Repositório Institucional da UNL
_version_ 1868414768319037440
author Baptista, Bruno Alexandre da Anunciação
author_facet Baptista, Bruno Alexandre da Anunciação
author_role author
contributor_name_str_mv Custódio, Ana
Brás, Carmo
RUN
country_str PT
creators_json_txt [{\"Person.name\":\"Baptista, Bruno Alexandre da Anunciação\"}]
datacite.contributors.contributor.contributorName.fl_str_mv Custódio, Ana
Brás, Carmo
RUN
datacite.creators.creator.creatorName.fl_str_mv Baptista, Bruno Alexandre da Anunciação
datacite.date.Accepted.fl_str_mv 2022-12-01T00:00:00Z
datacite.date.available.fl_str_mv 2023-11-15T11:57:02Z
datacite.date.embargoed.fl_str_mv 2023-11-15T11:57:02Z
datacite.rights.fl_str_mv http://purl.org/coar/access_right/c_abf2
datacite.subjects.subject.fl_str_mv Global optimization
Derivative-free optimization
Radial basis functions
Surrogate models
Direct search methods
Pattern search methods
datacite.titles.title.fl_str_mv INCORPORATING RADIAL BASIS FUNCTIONS IN GLOBAL AND LOCAL DIRECT SEARCH
dc.contributor.none.fl_str_mv Custódio, Ana
Brás, Carmo
RUN
dc.creator.none.fl_str_mv Baptista, Bruno Alexandre da Anunciação
dc.date.Accepted.fl_str_mv 2022-12-01T00:00:00Z
dc.date.available.fl_str_mv 2023-11-15T11:57:02Z
dc.date.embargoed.fl_str_mv 2023-11-15T11:57:02Z
dc.format.none.fl_str_mv application/pdf
dc.identifier.none.fl_str_mv http://hdl.handle.net/10362/159976
dc.language.none.fl_str_mv eng
dc.rights.none.fl_str_mv http://purl.org/coar/access_right/c_abf2
dc.subject.none.fl_str_mv Global optimization
Derivative-free optimization
Radial basis functions
Surrogate models
Direct search methods
Pattern search methods
dc.title.fl_str_mv INCORPORATING RADIAL BASIS FUNCTIONS IN GLOBAL AND LOCAL DIRECT SEARCH
dc.type.none.fl_str_mv http://purl.org/coar/resource_type/c_bdcc
description GLODS is a global derivative-free optimization algorithm, relying on local directional direct search, aided by a clever multistart strategy that does not conduct all the lines of search until the end. In 2015, time of the first release of the corresponding solver, GLODS was shown to be competitive when compared to state-of-the-art algorithms, such as MCS or DIRECT. GLODS resorts to sampling techniques to look for minima on a global scale, not taking advantage of the information gathered in previous iterations. As such, the main goal of this work is to replace the pseudo-random sampling approach, used by GLODS to initialize new lines of search, by the minimization of global models of the objective function, defined using radial basis functions, and computed using the points previously evaluated by the algorithm. This should allow a better placement of the starting points for new local lines of search, and, in turn, significantly increase the numerical performance of the algorithm. Naturally, incorporating radial basis functions in GLODS poses new challenges. In this work, we will address questions such as which radial basis functions to use, which points should be selected to compute them, how to minimize these functions, and how to take advantage of their minima in the execution of the algorithm. The new version of GLODS, incorporating radial basis functions, was calibrated to its best numerical performance, and then compared against other state-of-the-art solvers, such as MCS, DIRECT, MATSuMoTo, and ZOOpt. The results obtained are strongly positive. The new algorithm clearly outperforms its previous version, and is competitive with the other solvers tested. Finally, parallel strategies were implemented and tested. Results showed that it is very beneficial to evaluate multiple points simultaneously, for objective functions whose evaluation time is as low as 0.1 seconds. The proposed algorithm, called BoostGLODS, is a cutting-edge, powerful and efficient parallel global derivative-free optimization algo- rithm.
dirty 0
eu_rights_str_mv openAccess
format masterThesis
fulltext.url.fl_str_mv https://run.unl.pt/bitstreams/38ec10a4-dbf6-4dd7-bf24-640e922d014d/download
funding.funder.alternateName_str_mv FCT
funding.funder.identifier_str_mv http://doi.org/10.13039/501100001871
funding.funder.name_str_mv Fundação para a Ciência e a Tecnologia
funding.name_str_mv 3599-PPCDT
id run_71e6e665aa6ebe4e62d64f80e666fe4e
identifier.url.fl_str_mv http://hdl.handle.net/10362/159976
instacron_str unl
institution Universidade Nova de Lisboa
instname_str Universidade Nova de Lisboa
language eng
network_acronym_str run
network_name_str Repositório Institucional da UNL
oai_identifier_str oai:run.unl.pt:10362/159976
organization_str_mv urn:organizationAcronym:unl
person_str_mv Baptista, Bruno Alexandre da Anunciação
publishDate 2022
reponame_str Repositório Institucional da UNL
repository_id_str urn:repositoryAcronym:run
service_str_mv urn:repositoryAcronym:run
spelling engpt_PTGLODS is a global derivative-free optimization algorithm, relying on local directional direct search, aided by a clever multistart strategy that does not conduct all the lines of search until the end. In 2015, time of the first release of the corresponding solver, GLODS was shown to be competitive when compared to state-of-the-art algorithms, such as MCS or DIRECT. GLODS resorts to sampling techniques to look for minima on a global scale, not taking advantage of the information gathered in previous iterations. As such, the main goal of this work is to replace the pseudo-random sampling approach, used by GLODS to initialize new lines of search, by the minimization of global models of the objective function, defined using radial basis functions, and computed using the points previously evaluated by the algorithm. This should allow a better placement of the starting points for new local lines of search, and, in turn, significantly increase the numerical performance of the algorithm. Naturally, incorporating radial basis functions in GLODS poses new challenges. In this work, we will address questions such as which radial basis functions to use, which points should be selected to compute them, how to minimize these functions, and how to take advantage of their minima in the execution of the algorithm. The new version of GLODS, incorporating radial basis functions, was calibrated to its best numerical performance, and then compared against other state-of-the-art solvers, such as MCS, DIRECT, MATSuMoTo, and ZOOpt. The results obtained are strongly positive. The new algorithm clearly outperforms its previous version, and is competitive with the other solvers tested. Finally, parallel strategies were implemented and tested. Results showed that it is very beneficial to evaluate multiple points simultaneously, for objective functions whose evaluation time is as low as 0.1 seconds. The proposed algorithm, called BoostGLODS, is a cutting-edge, powerful and efficient parallel global derivative-free optimization algo- rithm.application/pdfpt_PTINCORPORATING RADIAL BASIS FUNCTIONS IN GLOBAL AND LOCAL DIRECT SEARCHBaptista, Bruno Alexandre da AnunciaçãoCustódio, AnaBrás, CarmoHostingInstitutionOrganizationalRUNe-mailmailto:run@unl.ptrun@unl.pt2023-11-15T11:57:02Z2022-122022-12-01T00:00:00ZHandlehttp://hdl.handle.net/10362/159976http://purl.org/coar/access_right/c_abf2open accessGlobal optimizationDerivative-free optimizationRadial basis functionsSurrogate modelsDirect search methodsPattern search methods4387452 bytesFundação para a Ciência e a TecnologiaImproving the performance and moving to newer dimensions in Derivative-Free Optimization3599-PPCDTCrossref Funder IDhttp://doi.org/10.13039/501100001871literaturehttp://purl.org/coar/resource_type/c_bdccmaster thesishttp://purl.org/coar/access_right/c_abf2application/pdffulltexthttps://run.unl.pt/bitstreams/38ec10a4-dbf6-4dd7-bf24-640e922d014d/download
spellingShingle INCORPORATING RADIAL BASIS FUNCTIONS IN GLOBAL AND LOCAL DIRECT SEARCH
Baptista, Bruno Alexandre da Anunciação
Global optimization
Derivative-free optimization
Radial basis functions
Surrogate models
Direct search methods
Pattern search methods
status SINGLETON
subject.fl_str_mv Global optimization
Derivative-free optimization
Radial basis functions
Surrogate models
Direct search methods
Pattern search methods
title INCORPORATING RADIAL BASIS FUNCTIONS IN GLOBAL AND LOCAL DIRECT SEARCH
title_full INCORPORATING RADIAL BASIS FUNCTIONS IN GLOBAL AND LOCAL DIRECT SEARCH
title_fullStr INCORPORATING RADIAL BASIS FUNCTIONS IN GLOBAL AND LOCAL DIRECT SEARCH
title_full_unstemmed INCORPORATING RADIAL BASIS FUNCTIONS IN GLOBAL AND LOCAL DIRECT SEARCH
title_short INCORPORATING RADIAL BASIS FUNCTIONS IN GLOBAL AND LOCAL DIRECT SEARCH
title_sort INCORPORATING RADIAL BASIS FUNCTIONS IN GLOBAL AND LOCAL DIRECT SEARCH
topic Global optimization
Derivative-free optimization
Radial basis functions
Surrogate models
Direct search methods
Pattern search methods
topic_facet Global optimization
Derivative-free optimization
Radial basis functions
Surrogate models
Direct search methods
Pattern search methods
url http://hdl.handle.net/10362/159976
visible 1