Automated Metaheuristic Algorithm Design With Autoregressive Learning

Automated design of metaheuristic algorithms offers an attractive avenue to reduce human effort and gain enhanced performance beyond human intuition. Current automated methods design algorithms within a fixed structure and operate from scratch. This poses a clear gap toward fully discovering potenti...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE transactions on evolutionary computation Jg. 29; H. 5; S. 2004 - 2018
Hauptverfasser: Zhao, Qi, Liu, Tengfei, Yan, Bai, Duan, Qiqi, Yang, Jian, Shi, Yuhui
Format: Journal Article
Sprache:Englisch
Veröffentlicht: IEEE 01.10.2025
Schlagworte:
ISSN:1089-778X, 1941-0026
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Abstract Automated design of metaheuristic algorithms offers an attractive avenue to reduce human effort and gain enhanced performance beyond human intuition. Current automated methods design algorithms within a fixed structure and operate from scratch. This poses a clear gap toward fully discovering potentials over the metaheuristic family and fertilizing from prior design experience. To bridge the gap, this article proposes an autoregressive learning-based designer for automated design of metaheuristic algorithms. Our designer formulates metaheuristic algorithm design as a sequence generation task, and harnesses an autoregressive generative network to handle the task. This offers two advances. First, through autoregressive inference, the designer generates algorithms with diverse lengths and structures, enabling to fully discover potentials over the metaheuristic family. Second, prior design knowledge learned and accumulated in neurons of the designer can be retrieved for designing algorithms for future problems, paving the way to continual design of algorithms for open-ended problem solving. Extensive experiments on numeral benchmarks and real-world problems reveal that the proposed designer generates algorithms that outperform all human-created baselines on 24 out of 25 test problems. The generated algorithms display various structures and behaviors, reasonably fitting for different problem-solving contexts. Code is available at https://github.com/auto4opt/ALDes .
AbstractList Automated design of metaheuristic algorithms offers an attractive avenue to reduce human effort and gain enhanced performance beyond human intuition. Current automated methods design algorithms within a fixed structure and operate from scratch. This poses a clear gap toward fully discovering potentials over the metaheuristic family and fertilizing from prior design experience. To bridge the gap, this article proposes an autoregressive learning-based designer for automated design of metaheuristic algorithms. Our designer formulates metaheuristic algorithm design as a sequence generation task, and harnesses an autoregressive generative network to handle the task. This offers two advances. First, through autoregressive inference, the designer generates algorithms with diverse lengths and structures, enabling to fully discover potentials over the metaheuristic family. Second, prior design knowledge learned and accumulated in neurons of the designer can be retrieved for designing algorithms for future problems, paving the way to continual design of algorithms for open-ended problem solving. Extensive experiments on numeral benchmarks and real-world problems reveal that the proposed designer generates algorithms that outperform all human-created baselines on 24 out of 25 test problems. The generated algorithms display various structures and behaviors, reasonably fitting for different problem-solving contexts. Code is available at https://github.com/auto4opt/ALDes .
Author Zhao, Qi
Shi, Yuhui
Liu, Tengfei
Duan, Qiqi
Yan, Bai
Yang, Jian
Author_xml – sequence: 1
  givenname: Qi
  orcidid: 0000-0003-4800-1136
  surname: Zhao
  fullname: Zhao, Qi
  email: zhaoq@sustech.edu.cn
  organization: Department of Computer Science and Engineering, Southern University of Science and Technology, Shenzhen, China
– sequence: 2
  givenname: Tengfei
  orcidid: 0009-0003-1435-9209
  surname: Liu
  fullname: Liu, Tengfei
  email: 12332470@mail.sustech.edu.cn
  organization: Department of Computer Science and Engineering, Southern University of Science and Technology, Shenzhen, China
– sequence: 3
  givenname: Bai
  orcidid: 0000-0003-3374-093X
  surname: Yan
  fullname: Yan, Bai
  email: yanb@sustech.edu.cn
  organization: Research Institute of Trustworthy Autonomous Systems, Southern University of Science and Technology, Shenzhen, China
– sequence: 4
  givenname: Qiqi
  orcidid: 0000-0002-4518-1701
  surname: Duan
  fullname: Duan, Qiqi
  email: 11749325@mail.sustech.edu.cn
  organization: Department of Computer Science and Engineering, Southern University of Science and Technology, Shenzhen, China
– sequence: 5
  givenname: Jian
  surname: Yang
  fullname: Yang, Jian
  email: yangj33@sustech.edu.cn
  organization: Department of Computer Science and Engineering, Southern University of Science and Technology, Shenzhen, China
– sequence: 6
  givenname: Yuhui
  orcidid: 0000-0002-8840-723X
  surname: Shi
  fullname: Shi, Yuhui
  email: shiyh@sustech.edu.cn
  organization: Department of Computer Science and Engineering, Southern University of Science and Technology, Shenzhen, China
BookMark eNpNkM1OwzAQhC1UJNrCAyBxyAuk7NpOYh-rUgpSEJfyc4scZ5MatQmyUyTenkTtgb3MjjQzh2_GJm3XEmO3CAtE0Pfb9ftqwYHLhZCpTLPsgk1RS4wBeDoZflA6zjL1ecVmIXwBoExQT9l6eey7g-mpil6oNzs6ehd6Z6Plvum863eH6IGCa9roYzDRmPbUeArB_VCUk_Gta5trdlmbfaCbs87Z2-N6u3qK89fN82qZx5aj6mNpQGgpqoQEotSlBVsJpTCtRZlUda2sAJ4gNxVPLJdlNpwC0KZEQtQk5gxPu9Z3IXiqi2_vDsb_FgjFyKEYORQjh-LMYejcnTqOiP7lUyUVJOIPZa1btw
CODEN ITEVF5
Cites_doi 10.1126/science.220.4598.671
10.1109/TEVC.2015.2474158
10.1145/2598394.2605681
10.1109/TETCI.2022.3146882
10.1126/sciadv.aap7885
10.48550/ARXIV.1706.03762
10.1145/3319619.3326913
10.1613/jair.1.13676
10.1007/978-3-319-96514-7
10.18653/v1/D16-1244
10.1162/evco_a_00263
10.1016/j.eswa.2022.119461
10.1038/s42256-022-00446-y
10.1016/0041-5553(67)90144-9
10.1109/MCI.2020.2976182
10.1109/TCOMM.2020.3033006
10.1007/3-540-48885-5_2
10.1109/CVPR.2018.00907
10.1016/j.ins.2013.04.015
10.1007/BF00175355
10.1109/ICEC.1998.699146
10.1287/ijoc.1.3.190
10.1016/j.asoc.2022.109725
10.1038/nature14236
10.1109/TEVC.2022.3197298
10.1007/978-3-642-01181-8_10
10.1007/0-306-48056-5_11
10.1109/JSAC.2020.3000813
10.1016/j.cor.2018.12.015
10.1162/evco.1993.1.1.1
10.1613/jair.2861
10.1016/j.eswa.2021.115493
10.1007/978-3-642-25566-3_40
10.1145/3205651.3208304
10.1073/pnas.1611835114
10.1109/TPAMI.2021.3057446
10.1007/978-3-319-91086-4_17
10.1016/j.orp.2016.09.002
10.1016/j.ejor.2021.05.042
10.1145/1543834.1543915
10.1016/S0305-0548(97)00031-2
10.1007/978-3-030-58115-2_10
10.1145/2001576.2001690
10.1057/jors.2013.71
10.1109/TEVC.2022.3159087
10.1007/978-3-642-17310-3_2
10.1109/TIT.2003.821988
10.1145/2464576.2482728
10.1109/CEC.2007.4424455
10.1007/978-3-030-72699-7_2
10.1109/TEVC.2021.3102863
10.3389/fphy.2014.00005
10.1038/s42256-021-00411-1
10.1109/SITIS.2010.22
10.1137/0202009
10.18653/v1/2023.acl-long.205
10.1162/evco_a_00240
10.1016/j.asoc.2019.106027
10.1162/evco_a_00242
10.1145/3319619.3326827
10.1007/978-3-319-30668-1_19
ContentType Journal Article
DBID 97E
RIA
RIE
AAYXX
CITATION
DOI 10.1109/TEVC.2024.3464677
DatabaseName IEEE All-Society Periodicals Package (ASPP) 2005–Present
IEEE All-Society Periodicals Package (ASPP) 1998–Present
IEEE/IET Electronic Library (IEL) (UW System Shared)
CrossRef
DatabaseTitle CrossRef
DatabaseTitleList
Database_xml – sequence: 1
  dbid: RIE
  name: IEEE/IET Electronic Library (IEL) (UW System Shared)
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
DeliveryMethod fulltext_linktorsrc
Discipline Engineering
Computer Science
EISSN 1941-0026
EndPage 2018
ExternalDocumentID 10_1109_TEVC_2024_3464677
10684805
Genre orig-research
GrantInformation_xml – fundername: Shenzhen Municipal Fundamental Research Program; Shenzhen Fundamental Research Program
  grantid: JCYJ20200109141235597
  funderid: 10.13039/501100017607
– fundername: Program for Guangdong Introducing Innovative and Entrepreneurial Teams
  grantid: 2017ZT07X386
– fundername: Guangdong Basic and Applied Basic Research Foundation
  grantid: 2024A1515012241; 2021A1515110024
  funderid: 10.13039/501100001321
– fundername: National Science Foundation of China
  grantid: 72401122; 61761136008
  funderid: 10.13039/501100001809
– fundername: Shenzhen Peacock Plan
  grantid: KQTD2016112514355531
  funderid: 10.13039/501100012234
GroupedDBID -~X
.DC
0R~
29I
4.4
5GY
5VS
6IF
6IK
97E
AAJGR
AASAJ
AAWTH
ABAZT
ABJNI
ABQJQ
ABVLG
ACGFO
ACGFS
ACIWK
AENEX
AETIX
AGQYO
AGSQL
AHBIQ
AI.
AIBXA
AKJIK
AKQYR
ALLEH
ALMA_UNASSIGNED_HOLDINGS
ATWAV
BEFXN
BFFAM
BGNUA
BKEBE
BPEOZ
CS3
EBS
EJD
HZ~
H~9
IFIPE
IFJZH
IPLJI
JAVBF
LAI
M43
O9-
OCL
P2P
PQQKQ
RIA
RIE
RNS
TN5
VH1
AAYXX
CITATION
ID FETCH-LOGICAL-c218t-4a03943d5e31149bc0cd38816f3b5dff8c302512ad25c24b77778009ab1e119e3
IEDL.DBID RIE
ISICitedReferencesCount 1
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001591697300026&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 1089-778X
IngestDate Sat Nov 29 07:12:18 EST 2025
Sat Oct 25 03:12:18 EDT 2025
IsPeerReviewed true
IsScholarly true
Issue 5
Language English
License https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html
https://doi.org/10.15223/policy-029
https://doi.org/10.15223/policy-037
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-c218t-4a03943d5e31149bc0cd38816f3b5dff8c302512ad25c24b77778009ab1e119e3
ORCID 0009-0003-1435-9209
0000-0002-4518-1701
0000-0003-4800-1136
0000-0002-8840-723X
0000-0003-3374-093X
PageCount 15
ParticipantIDs crossref_primary_10_1109_TEVC_2024_3464677
ieee_primary_10684805
PublicationCentury 2000
PublicationDate 2025-10-01
PublicationDateYYYYMMDD 2025-10-01
PublicationDate_xml – month: 10
  year: 2025
  text: 2025-10-01
  day: 01
PublicationDecade 2020
PublicationTitle IEEE transactions on evolutionary computation
PublicationTitleAbbrev TEVC
PublicationYear 2025
Publisher IEEE
Publisher_xml – name: IEEE
References ref13
ref57
ref12
ref56
ref15
ref59
ref14
ref58
ref11
ref55
ref10
ref54
ref17
ref16
ref19
ref18
ref50
Glover (ref1) 2006
ref48
ref41
ref44
ref43
Kingma (ref64) 2017
ref49
ref8
ref7
ref4
ref6
ref5
Schulman (ref52)
ref40
Loynd (ref51)
Goodfellow (ref46) 2016
ref35
ref34
ref37
ref36
ref31
ref30
ref33
ref32
ref2
ref39
Mikolov (ref53) 2013
Zhao (ref3) 2024
ref71
Zoph (ref47)
ref70
ref73
ref72
ref24
ref68
ref23
ref67
ref26
ref25
ref69
ref20
ref63
ref22
ref66
Lindauer (ref9) 2022; 23
ref21
ref65
ref28
ref27
Adriaensen (ref38)
ref29
ref60
ref62
ref61
Schulman (ref42) 2017
Lin (ref45) 2017
References_xml – ident: ref32
  doi: 10.1126/science.220.4598.671
– ident: ref16
  doi: 10.1109/TEVC.2015.2474158
– ident: ref37
  doi: 10.1145/2598394.2605681
– ident: ref21
  doi: 10.1109/TETCI.2022.3146882
– ident: ref49
  doi: 10.1126/sciadv.aap7885
– ident: ref43
  doi: 10.48550/ARXIV.1706.03762
– ident: ref72
  doi: 10.1145/3319619.3326913
– ident: ref7
  doi: 10.1613/jair.1.13676
– ident: ref13
  doi: 10.1007/978-3-319-96514-7
– ident: ref44
  doi: 10.18653/v1/D16-1244
– ident: ref18
  doi: 10.1162/evco_a_00263
– ident: ref40
  doi: 10.1016/j.eswa.2022.119461
– year: 2017
  ident: ref64
  article-title: Adam: A method for stochastic optimization
  publication-title: arXiv:1412.6980
– ident: ref67
  doi: 10.1038/s42256-022-00446-y
– ident: ref56
  doi: 10.1016/0041-5553(67)90144-9
– ident: ref5
  doi: 10.1109/MCI.2020.2976182
– ident: ref68
  doi: 10.1109/TCOMM.2020.3033006
– ident: ref29
  doi: 10.1007/3-540-48885-5_2
– start-page: 1889
  volume-title: Proc. Int. Conf. Mach. Learn.
  ident: ref52
  article-title: Trust region policy optimization
– ident: ref48
  doi: 10.1109/CVPR.2018.00907
– ident: ref58
  doi: 10.1016/j.ins.2013.04.015
– start-page: 554
  volume-title: Proc. 25th Int. Joint Conf. Artif. Intell.
  ident: ref38
  article-title: Towards a white box approach to automated algorithm design
– ident: ref14
  doi: 10.1007/BF00175355
– ident: ref33
  doi: 10.1109/ICEC.1998.699146
– start-page: 6404
  volume-title: Proc. Int. Conf. Mach. Learn.
  ident: ref51
  article-title: Working memory graphs
– ident: ref63
  doi: 10.1287/ijoc.1.3.190
– ident: ref69
  doi: 10.1016/j.asoc.2022.109725
– ident: ref41
  doi: 10.1038/nature14236
– ident: ref20
  doi: 10.1109/TEVC.2022.3197298
– ident: ref30
  doi: 10.1007/978-3-642-01181-8_10
– ident: ref62
  doi: 10.1007/0-306-48056-5_11
– ident: ref71
  doi: 10.1109/JSAC.2020.3000813
– ident: ref35
  doi: 10.1016/j.cor.2018.12.015
– start-page: 1
  year: 2024
  ident: ref3
  article-title: Automated design of metaheuristic algorithms: A survey
  publication-title: Trans. Mach. Learn. Res.
– ident: ref15
  doi: 10.1162/evco.1993.1.1.1
– start-page: 1
  volume-title: Proc. Int. Conf. Learn. Represent.
  ident: ref47
  article-title: Neural architecture search with reinforcement learning
– ident: ref11
  doi: 10.1613/jair.2861
– ident: ref39
  doi: 10.1016/j.eswa.2021.115493
– ident: ref8
  doi: 10.1007/978-3-642-25566-3_40
– ident: ref23
  doi: 10.1145/3205651.3208304
– ident: ref59
  doi: 10.1073/pnas.1611835114
– year: 2017
  ident: ref42
  article-title: Proximal policy optimization algorithms
  publication-title: arXiv:1707.06347
– ident: ref60
  doi: 10.1109/TPAMI.2021.3057446
– volume: 23
  start-page: 1
  issue: 54
  year: 2022
  ident: ref9
  article-title: SMAC3: A versatile Bayesian optimization package for hyperparameter optimization
  publication-title: J. Mach. Learn. Res.
– ident: ref4
  doi: 10.1007/978-3-319-91086-4_17
– ident: ref10
  doi: 10.1016/j.orp.2016.09.002
– ident: ref2
  doi: 10.1016/j.ejor.2021.05.042
– ident: ref25
  doi: 10.1145/1543834.1543915
– year: 2017
  ident: ref45
  article-title: A structured self-attentive sentence embedding
  publication-title: arXiv:1703.03130
– volume-title: Deep Learning
  year: 2016
  ident: ref46
– ident: ref65
  doi: 10.1016/S0305-0548(97)00031-2
– ident: ref57
  doi: 10.1007/978-3-030-58115-2_10
– ident: ref55
  doi: 10.1145/2001576.2001690
– volume-title: Handbook of Metaheuristics
  year: 2006
  ident: ref1
– ident: ref12
  doi: 10.1057/jors.2013.71
– ident: ref22
  doi: 10.1109/TEVC.2022.3159087
– ident: ref28
  doi: 10.1007/978-3-642-17310-3_2
– ident: ref70
  doi: 10.1109/TIT.2003.821988
– ident: ref27
  doi: 10.1145/2464576.2482728
– ident: ref26
  doi: 10.1109/CEC.2007.4424455
– ident: ref73
  doi: 10.1007/978-3-030-72699-7_2
– ident: ref19
  doi: 10.1109/TEVC.2021.3102863
– ident: ref66
  doi: 10.3389/fphy.2014.00005
– ident: ref50
  doi: 10.1038/s42256-021-00411-1
– ident: ref36
  doi: 10.1109/SITIS.2010.22
– ident: ref34
  doi: 10.1137/0202009
– ident: ref54
  doi: 10.18653/v1/2023.acl-long.205
– year: 2013
  ident: ref53
  article-title: Efficient estimation of word representations in vector space
  publication-title: arXiv:1301.3781
– ident: ref17
  doi: 10.1162/evco_a_00240
– ident: ref61
  doi: 10.1016/j.asoc.2019.106027
– ident: ref6
  doi: 10.1162/evco_a_00242
– ident: ref24
  doi: 10.1145/3319619.3326827
– ident: ref31
  doi: 10.1007/978-3-319-30668-1_19
SSID ssj0014519
Score 2.4929075
Snippet Automated design of metaheuristic algorithms offers an attractive avenue to reduce human effort and gain enhanced performance beyond human intuition. Current...
SourceID crossref
ieee
SourceType Index Database
Publisher
StartPage 2004
SubjectTerms Approximation algorithms
Automated design
evolutionary algorithm
learning
metaheuristic
Metaheuristics
Neurons
optimization
Problem-solving
Search problems
transformer
Transformers
Vectors
Title Automated Metaheuristic Algorithm Design With Autoregressive Learning
URI https://ieeexplore.ieee.org/document/10684805
Volume 29
WOSCitedRecordID wos001591697300026&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
journalDatabaseRights – providerCode: PRVIEE
  databaseName: IEEE/IET Electronic Library (IEL) (UW System Shared)
  customDbUrl:
  eissn: 1941-0026
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0014519
  issn: 1089-778X
  databaseCode: RIE
  dateStart: 19970101
  isFulltext: true
  titleUrlDefault: https://ieeexplore.ieee.org/
  providerName: IEEE
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV09T8MwELWgYoCBQimifMkDE1JKEjv-GKvSioWKoUC3KLYvbSVoUUj4_dhOisrAQKYk8mC9xL473917CN0kQipGExPkmoYBFYYFmQIIVBIrrg1jRudebIJPJmI2k09Ns7rvhQEAX3wGfXfrc_lmrSt3VGZXOBNUOMbSXc5Z3az1kzJwPCl1Nb20LqOYNSnMKJR309HL0IaCMe0TyuzOwH8ZoS1VFW9Uxu1_TucIHTbeIx7Un_sY7cCqg9obZQbcLNQOOtiiGTxBo0FVrq1nCgY_QpktoKrpmfHgbb4uluXiHd_7Qg78ah-wG12Aj8PtVogbBtZ5Fz2PR9PhQ9DIJwTa2u0yoFlIJCUmAWKDHql0qA0RImI5UYnJc6GJCzDizMSJjqni9rLuo8xUBFEkgZyi1mq9gjOEaU5NbiCJNFGUcimcSIlUMXApNVOkh243eKYfNUtG6qOLUKYO_NSBnzbg91DXYbk1sIbx_I_3F2g_dqK7voLuErXKooIrtKe_yuVnce1_gm9VDLAZ
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3PT8IwFG4MmqgHUcSIP3vwZDLc1m5rjwQhGIF4QOW2rO0bkCiYufn323bD4MGDO21LD823te-9vve-D6GbgHER0kA5qaSuQ5kKnUQAOCLwRSRVGCqZWrGJaDxm0yl_qprVbS8MANjiM2ibW5vLVytZmKMyvcJDRplhLN0OKPXdsl3rJ2lgmFLKenqunUY2rZKYnsvvJr2Xrg4GfdomNNR7Q_TLDG3oqliz0q__c0KH6KDyH3Gn_OBHaAuWDVRfazPgaqk20P4G0eAx6nWKfKV9U1B4BHkyh6IkaMadt9kqW-Tzd3xvSznwq37AZnQGNhLXmyGuOFhnTfTc7026A6cSUHCktty5QxOXcEpUAESHPVxIVyrCmBemRAQqTZkkJsTwE-UH0qci0pd2IHkiPPA8DuQE1ZarJZwiTFOqUgWBJ4mgNOLMyJRw4UPEuQwFaaHbNZ7xR8mTEdv4wuWxAT824McV-C3UNFhuDCxhPPvj_TXaHUxGw3j4MH48R3u-keC19XQXqJZnBVyiHfmVLz6zK_tDfAOu4rNg
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Automated+Metaheuristic+Algorithm+Design+With+Autoregressive+Learning&rft.jtitle=IEEE+transactions+on+evolutionary+computation&rft.au=Zhao%2C+Qi&rft.au=Liu%2C+Tengfei&rft.au=Yan%2C+Bai&rft.au=Duan%2C+Qiqi&rft.date=2025-10-01&rft.issn=1089-778X&rft.eissn=1941-0026&rft.volume=29&rft.issue=5&rft.spage=2004&rft.epage=2018&rft_id=info:doi/10.1109%2FTEVC.2024.3464677&rft.externalDBID=n%2Fa&rft.externalDocID=10_1109_TEVC_2024_3464677
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1089-778X&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1089-778X&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1089-778X&client=summon