Learning Value Functions in Interactive Evolutionary Multiobjective Optimization

This paper proposes an interactive multiobjective evolutionary algorithm (MOEA) that attempts to learn a value function capturing the users' true preferences. At regular intervals, the user is asked to rank a single pair of solutions. This information is used to update the algorithm's inte...

Full description

Saved in:
Bibliographic Details
Published inIEEE transactions on evolutionary computation Vol. 19; no. 1; pp. 88 - 102
Main Authors Branke, Jurgen, Greco, Salvatore, Slowinski, Roman, Zielniewicz, Piotr
Format Journal Article
LanguageEnglish
Published IEEE 01.02.2015
Subjects
Additives
Business
Computational modeling
Educational institutions
Electronic mail
Linear programming
Optimization
ordinal regression
interactive procedure
Evolutionary multiobjective optimization
preference learning
Online AccessGet full text

Cover

Abstract This paper proposes an interactive multiobjective evolutionary algorithm (MOEA) that attempts to learn a value function capturing the users' true preferences. At regular intervals, the user is asked to rank a single pair of solutions. This information is used to update the algorithm's internal value function model, and the model is used in subsequent generations to rank solutions incomparable according to dominance. This speeds up evolution toward the region of the Pareto front that is most desirable to the user. We take into account the most general additive value function as a preference model and we empirically compare different ways to identify the value function that seems to be the most representative with respect to the given preference information, different types of user preferences, and different ways to use the learned value function in the MOEA. Results on a number of different scenarios suggest that the proposed algorithm works well over a range of benchmark problems and types of user preferences.
AbstractList This paper proposes an interactive multiobjective evolutionary algorithm (MOEA) that attempts to learn a value function capturing the users' true preferences. At regular intervals, the user is asked to rank a single pair of solutions. This information is used to update the algorithm's internal value function model, and the model is used in subsequent generations to rank solutions incomparable according to dominance. This speeds up evolution toward the region of the Pareto front that is most desirable to the user. We take into account the most general additive value function as a preference model and we empirically compare different ways to identify the value function that seems to be the most representative with respect to the given preference information, different types of user preferences, and different ways to use the learned value function in the MOEA. Results on a number of different scenarios suggest that the proposed algorithm works well over a range of benchmark problems and types of user preferences.
Author Greco, Salvatore
Branke, Jurgen
Zielniewicz, Piotr
Slowinski, Roman
Author_xml – sequence: 1
  givenname: Jurgen
  surname: Branke
  fullname: Branke, Jurgen
  email: juergen.branke@wbs.ac.uk
  organization: Bus. Sch., Univ. of Warwick, Coventry, UK
– sequence: 2
  givenname: Salvatore
  surname: Greco
  fullname: Greco, Salvatore
  email: salgreco@unict.it
  organization: Dept. of Econ. & Bus., Univ. of Catania, Catania, Italy
– sequence: 3
  givenname: Roman
  surname: Slowinski
  fullname: Slowinski, Roman
  email: roman.slowinski@cs.put.poznan.pl
  organization: Inst. of Comput. Sci., Poznan Univ. of Technol., Poznan, Poland
– sequence: 4
  givenname: Piotr
  surname: Zielniewicz
  fullname: Zielniewicz, Piotr
  email: piotr.zielniewicz@cs.put.poznan.pl
  organization: Inst. of Comput. Sci., Poznan Univ. of Technol., Poznan, Poland
BookMark eNp9UEFOwzAQtFCRaAsPQFzygZR17CT2EVUtrVRUDqXiFtnuBrlKncpJKsHrcWjFgQOnndXMrHZmRAaudkjIPYUJpSAfN7PtdJIA5ZOEAcsFuyJDKjmNAZJsEDAIGee5eL8ho6bZQ1CmVA7J6wqVd9Z9RFtVdRjNO2daW7smsi5auha9CvsJo9mprrqeUf4zeumqAPUez9z62NqD_VI9fUuuS1U1eHeZY_I2n22mi3i1fl5On1axYYy2cSZBl0Ki5pgxjkabhHPBldwhM5ornUoVBCoTCFwDU2nGFYgSEyh3ISAbk_x81_i6aTyWhbHtzwetV7YqKBR9MUVfTNEXU1yKCU76x3n09hBi_et5OHssIv7qszyRkKbsG0gccyA
CODEN ITEVF5
CitedBy_id crossref_primary_10_1109_TCYB_2018_2859363
crossref_primary_10_1016_j_ejor_2015_10_027
crossref_primary_10_1109_TAI_2024_3444736
crossref_primary_10_1109_TNNLS_2018_2847412
crossref_primary_10_1016_j_eswa_2020_113563
crossref_primary_10_1109_TFUZZ_2018_2880700
crossref_primary_10_1109_TEVC_2015_2512930
crossref_primary_10_1016_j_omega_2017_06_013
crossref_primary_10_1016_j_ins_2018_06_034
crossref_primary_10_1109_TEVC_2020_2987559
crossref_primary_10_3390_mca27060112
crossref_primary_10_1109_TEVC_2019_2915767
crossref_primary_10_1016_j_omega_2016_07_003
crossref_primary_10_1109_MCI_2015_2405275
crossref_primary_10_1016_j_swevo_2019_100602
crossref_primary_10_1007_s10726_016_9506_6
crossref_primary_10_1016_j_procs_2020_08_030
crossref_primary_10_1145_3448301
crossref_primary_10_1145_3582004
crossref_primary_10_1016_j_ejor_2019_11_023
crossref_primary_10_1016_j_swevo_2018_04_009
crossref_primary_10_1142_S0219622020500510
crossref_primary_10_1016_j_cor_2019_04_008
crossref_primary_10_1109_ACCESS_2018_2856832
crossref_primary_10_1016_j_ins_2022_10_064
crossref_primary_10_1016_j_asoc_2024_111950
crossref_primary_10_1109_TEVC_2016_2605501
crossref_primary_10_1108_K_07_2024_1780
crossref_primary_10_1016_j_trb_2025_103162
crossref_primary_10_1007_s11831_021_09562_1
crossref_primary_10_1007_s10726_024_09915_8
crossref_primary_10_1016_j_omega_2022_102715
crossref_primary_10_3846_tede_2019_9475
crossref_primary_10_1016_j_cie_2023_109546
crossref_primary_10_3390_technologies10010024
crossref_primary_10_1016_j_omega_2024_103203
crossref_primary_10_1007_s11047_018_9685_y
crossref_primary_10_1038_nature14544
crossref_primary_10_20965_jaciii_2017_p0284
crossref_primary_10_1016_j_cie_2022_108022
crossref_primary_10_1109_TEVC_2018_2865495
crossref_primary_10_1016_j_ejor_2023_01_062
crossref_primary_10_1109_TEVC_2019_2925175
crossref_primary_10_1016_j_ins_2020_11_030
crossref_primary_10_1631_jzus_A1600034
crossref_primary_10_1016_j_omega_2023_102925
crossref_primary_10_1109_TEVC_2016_2521868
crossref_primary_10_1002_mcda_1599
crossref_primary_10_1007_s11573_015_0801_5
crossref_primary_10_1109_TEVC_2023_3234269
crossref_primary_10_1016_j_knosys_2023_110871
crossref_primary_10_1007_s00500_016_2099_9
crossref_primary_10_1016_j_asoc_2020_106432
crossref_primary_10_1109_ACCESS_2018_2797879
crossref_primary_10_1016_j_ejdp_2024_100047
Cites_doi 10.1016/S0965-9978(00)00110-1
10.1287/mnsc.49.12.1726.25117
10.1007/0-387-23081-5_8
10.1007/978-3-642-01020-0_43
10.1145/2001576.2001675
10.1109/CEC.2002.1007032
10.1016/j.ejor.2008.02.006
10.2478/v10175-010-0033-3
10.1016/j.asoc.2009.08.019
10.1016/S0377-2217(00)00167-3
10.1109/4235.996017
10.1109/CEC.2000.870272
10.1016/S0377-2217(00)00042-4
10.1007/978-1-4757-5184-0
10.1109/TEVC.2005.861417
10.1007/978-3-642-13800-3_35
10.1007/978-1-4419-5904-1_9
10.1287/mnsc.30.11.1336
10.1007/BF02291490
10.1007/978-3-540-88908-3
10.1287/mnsc.20.8.1217
10.1109/GEFS.2008.4484566
10.1109/TEVC.2010.2064323
10.1007/978-3-540-88908-3_6
10.1109/TEVC.2010.2070070
10.1016/0377-2217(82)90155-2
10.1007/b100605
10.1109/TEVC.2010.2058119
10.1016/j.ejor.2010.02.027
10.1016/j.ejor.2007.08.013
10.1109/TEVC.2010.2041060
10.2307/3003600
10.1007/s001580050111
10.1109/TEVC.2010.2058118
10.1007/978-3-642-37140-0_10
10.1162/106365600568202
10.1023/A:1008210427517
10.1016/0167-9236(94)90049-3
10.1109/TSMC.1984.6313205
ContentType Journal Article
DBID 97E
RIA
RIE
AAYXX
CITATION
DOI 10.1109/TEVC.2014.2303783
DatabaseName IEEE All-Society Periodicals Package (ASPP) 2005–Present
IEEE All-Society Periodicals Package (ASPP) 1998–Present
IEEE Electronic Library (IEL)
CrossRef
DatabaseTitle CrossRef
DatabaseTitleList
Database_xml – sequence: 1
  dbid: RIE
  name: IEEE Electronic Library (IEL)
  url: https://proxy.k.utb.cz/login?url=https://ieeexplore.ieee.org/
  sourceTypes: Publisher
DeliveryMethod fulltext_linktorsrc
Discipline Engineering
Computer Science
Business
EISSN 1941-0026
EndPage 102
ExternalDocumentID 10_1109_TEVC_2014_2303783
6729055
Genre orig-research
GrantInformation_xml – fundername: Polish National Science Center
  grantid: DEC-2011/01/B/ST6/07318
GroupedDBID -~X
.DC
0R~
29I
4.4
5GY
5VS
6IF
6IK
6IL
6IN
97E
AAJGR
AARMG
AASAJ
AAWTH
ABAZT
ABJNI
ABQJQ
ABVLG
ACGFO
ACGFS
ACIWK
ADZIZ
AENEX
AETIX
AGQYO
AGSQL
AHBIQ
AI.
AIBXA
AKJIK
AKQYR
ALLEH
ALMA_UNASSIGNED_HOLDINGS
ATWAV
BEFXN
BFFAM
BGNUA
BKEBE
BPEOZ
CHZPO
CS3
EBS
EJD
HZ~
H~9
IEGSK
IFIPE
IFJZH
IPLJI
JAVBF
LAI
M43
O9-
OCL
P2P
PQQKQ
RIA
RIE
RIL
RNS
TN5
VH1
AAYOK
AAYXX
CITATION
RIG
ID FETCH-LOGICAL-c331t-690bf89eb4e634ecbc24484a9de3cb4ab59abf8a68e04b03a564a08fe20fd0373
IEDL.DBID RIE
ISSN 1089-778X
IngestDate Tue Jul 01 01:56:20 EDT 2025
Thu Apr 24 23:07:32 EDT 2025
Tue Aug 26 16:40:05 EDT 2025
IsPeerReviewed true
IsScholarly true
Issue 1
Keywords ordinal regression
interactive procedure
Evolutionary multiobjective optimization
preference learning
Language English
License https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-c331t-690bf89eb4e634ecbc24484a9de3cb4ab59abf8a68e04b03a564a08fe20fd0373
PageCount 15
ParticipantIDs crossref_citationtrail_10_1109_TEVC_2014_2303783
ieee_primary_6729055
crossref_primary_10_1109_TEVC_2014_2303783
ProviderPackageCode CITATION
AAYXX
PublicationCentury 2000
PublicationDate 2015-Feb.
2015-2-00
PublicationDateYYYYMMDD 2015-02-01
PublicationDate_xml – month: 02
  year: 2015
  text: 2015-Feb.
PublicationDecade 2010
PublicationTitle IEEE transactions on evolutionary computation
PublicationTitleAbbrev TEVC
PublicationYear 2015
Publisher IEEE
Publisher_xml – name: IEEE
References ref13
ref12
ref15
ref14
ref52
ref11
ref10
ref17
ref16
ref19
ref18
ref51
ref50
ref46
ref45
barbosa (ref37) 2001
ref48
ref47
ref41
ref43
rachmawati (ref35) 2006
ref49
ref7
ref9
ref6
branke (ref5) 2008
ref40
steuer (ref3) 1986
fonseca (ref23) 1993
ref34
ref36
ref31
ref30
ref33
ref32
ref1
jaimes (ref24) 2011
todd (ref39) 1999
ref38
jaszkiewicz (ref42) 2007; 32
branke (ref44) 2004
greco (ref22) 2011
keeney (ref20) 1976
figueira (ref8) 2005
ref26
ref25
miettinen (ref2) 1999
ref21
vincke (ref4) 1992
ref27
ref29
greenwood (ref28) 1997
References_xml – ident: ref26
  doi: 10.1016/S0965-9978(00)00110-1
– ident: ref36
  doi: 10.1287/mnsc.49.12.1726.25117
– ident: ref13
  doi: 10.1007/0-387-23081-5_8
– ident: ref6
  doi: 10.1007/978-3-642-01020-0_43
– year: 1992
  ident: ref4
  publication-title: Multicriteria Decision Aid
– ident: ref31
  doi: 10.1145/2001576.2001675
– ident: ref51
  doi: 10.1109/CEC.2002.1007032
– ident: ref17
  doi: 10.1016/j.ejor.2008.02.006
– volume: 32
  start-page: 15
  year: 2007
  ident: ref42
  article-title: Interactive multiobjective optimization with the Pareto memetic algorithm
  publication-title: Found Comput Decision Sci
– ident: ref7
  doi: 10.2478/v10175-010-0033-3
– ident: ref29
  doi: 10.1016/j.asoc.2009.08.019
– start-page: 1605
  year: 2011
  ident: ref24
  article-title: Preference incorporation to solve many-ojective airfoil design problems
  publication-title: Proc Congr Evol Comput
– ident: ref16
  doi: 10.1016/S0377-2217(00)00167-3
– year: 1999
  ident: ref2
  publication-title: Nonlinear Multiobjective Optimization
– ident: ref49
  doi: 10.1109/4235.996017
– ident: ref34
  doi: 10.1109/CEC.2000.870272
– ident: ref21
  doi: 10.1016/S0377-2217(00)00042-4
– ident: ref33
  doi: 10.1007/978-1-4757-5184-0
– start-page: 1738
  year: 1999
  ident: ref39
  article-title: Directed multiple objective search of design spaces using genetic algorithms and neural networks
  publication-title: Proc Genetic Evol Comput Conf
– year: 1976
  ident: ref20
  publication-title: Decisions With Multiple Objectives Preferences and Value Tradeoffs
– ident: ref52
  doi: 10.1109/TEVC.2005.861417
– ident: ref41
  doi: 10.1007/978-3-642-13800-3_35
– ident: ref19
  doi: 10.1007/978-1-4419-5904-1_9
– ident: ref47
  doi: 10.1287/mnsc.30.11.1336
– ident: ref10
  doi: 10.1007/BF02291490
– year: 1986
  ident: ref3
  publication-title: Multiple Criteria Optimization Theory Computation and Application
– year: 2008
  ident: ref5
  publication-title: Multiobjective Optimization Interactive and Evolutionary Approaches
  doi: 10.1007/978-3-540-88908-3
– ident: ref11
  doi: 10.1287/mnsc.20.8.1217
– ident: ref1
  doi: 10.1109/GEFS.2008.4484566
– ident: ref38
  doi: 10.1109/TEVC.2010.2064323
– ident: ref32
  doi: 10.1007/978-3-540-88908-3_6
– ident: ref30
  doi: 10.1109/TEVC.2010.2070070
– year: 2011
  ident: ref22
  article-title: Parsimonious preference models for robust ordinal regression
  publication-title: Proc EWG-MCDA
– start-page: 416
  year: 1993
  ident: ref23
  article-title: Genetic algorithms for multiobjective optimization: Formulation, discussion, and generalization
  publication-title: Proc 5th Int Conf Genetic Algorithms
– ident: ref12
  doi: 10.1016/0377-2217(82)90155-2
– year: 2005
  ident: ref8
  publication-title: Multiple Criteria Decision Analysis State of the Art Surveys
  doi: 10.1007/b100605
– ident: ref27
  doi: 10.1109/TEVC.2010.2058119
– ident: ref48
  doi: 10.1016/j.ejor.2010.02.027
– ident: ref18
  doi: 10.1016/j.ejor.2007.08.013
– ident: ref25
  doi: 10.1109/TEVC.2010.2041060
– start-page: 437
  year: 1997
  ident: ref28
  article-title: Fitness functions for multiple objective optimization problems: Combining preferences with Pareto rankings
  publication-title: Foundations of Genetic Algorithms
– ident: ref9
  doi: 10.2307/3003600
– start-page: 722
  year: 2004
  ident: ref44
  article-title: Finding knees in multi-objective optimization
  publication-title: Proc PPSN
– ident: ref45
  doi: 10.1007/s001580050111
– start-page: 203
  year: 2001
  ident: ref37
  article-title: An interactive genetic algorithm with co-evolution of weights for multiobjective problems
  publication-title: Proc Genetic Evol Comput Conf
– ident: ref40
  doi: 10.1109/TEVC.2010.2058118
– ident: ref46
  doi: 10.1007/978-3-642-37140-0_10
– ident: ref50
  doi: 10.1162/106365600568202
– ident: ref15
  doi: 10.1023/A:1008210427517
– ident: ref14
  doi: 10.1016/0167-9236(94)90049-3
– start-page: 3385
  year: 2006
  ident: ref35
  article-title: Preference incorporation in multiobjective evolutionary algorithms: A survey
  publication-title: Proc Congr Evol Comput
– ident: ref43
  doi: 10.1109/TSMC.1984.6313205
SSID ssj0014519
Score 2.449448
Snippet This paper proposes an interactive multiobjective evolutionary algorithm (MOEA) that attempts to learn a value function capturing the users' true preferences....
SourceID crossref
ieee
SourceType Enrichment Source
Index Database
Publisher
StartPage 88
SubjectTerms Additives
Business
Computational modeling
Educational institutions
Electronic mail
Linear programming
Optimization
Title Learning Value Functions in Interactive Evolutionary Multiobjective Optimization
URI https://ieeexplore.ieee.org/document/6729055
Volume 19
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
link http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwjV1LSwMxEB5sD6IHH63imxw8idvGTXa7OUppKYKPQ1t6W5LsrPhqRVpBf71JNrtUEfG2LBMIO7PzSL75BuCUhcYrah4GkvFOwEPJAhUzU_OgiQ05irCD9rzj-iYejPjVJJqswHnVC4OIDnyGLfvo7vKzmV7Yo7J2bDJBGkU1qJnCrejVqm4MLE1KAaYXJmNMJv4G84KK9rA37loQF7eoZ9ZJ2LcYtDRUxcWU_iZcl7spoCRPrcVctfTnD6LG_253CzZ8ckkuC2vYhhWcNmC1xLY3YLOc4UD8L92A9SVCwibcebrVezKWzwskfRP1nGGShylxh4fS-UfSe_cmK98-iGvinanHwneSW-OFXnx75w6M-r1hdxD4mQuBZuxiHphiWeWJQMUxZhy10ib-J1yKDJlWXKpISCMg4wQpV5TJKOaSJjmGNM_Mx2W7UJ_OprgHBLXscMlMxomay8zyyiuKochDzFioxD7QUgup9oTkdi7Gc-oKEypSq7jUKi71ituHs2rJa8HG8Zdw0-qkEvTqOPj99SGsmcVRAcg-gvr8bYHHJt-YqxNnaF8adNJu
linkProvider IEEE
linkToHtml http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwjV1LT8MwDLZ4SDwOPAaINzlwQnQLSdo1R4Q2jceAw0C7VUnqIl4bQhsS_HqSNJsAIcStqtwqql3bcT5_BtjnzHpFI1ikuKhHgike6YTbPQ_a2FCgZHV09Y72ZdK6EWfduDsBh-NeGET04DOsukt_lp_3zdCVymqJzQRpHE_CtI37MSu7tcZnBo4opYTTS5szpt1whnlEZa3TuD1xMC7hcM-8nvJvUejLWBUfVZqL0B6tpwSTPFaHA101Hz-oGv-74CVYCOklOS7tYRkmsFeBmRG6vQKLoykOJPzUFZj_Qkm4AteBcPWO3KqnIZKmjXveNMl9j_jyofIekjTegtGq13fi23j7-qH0nuTK-qHn0OC5CjfNRuekFYWpC5Hh_GgQ2e2yLlKJWmDCBRptbAaQCiVz5EYLpWOprIBKUqRCU67iRCiaFshokduPy9dgqtfv4ToQNKouFLc5JxqhcscsrykyWTDMOdNyA-hIC5kJlORuMsZT5rcmVGZOcZlTXBYUtwEH40deSj6Ov4RXnE7GgkEdm7_f3oPZVqd9kV2cXp5vwZx9UVzCs7dhavA6xB2bfQz0rje6Tz-41bg
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=Learning+Value+Functions+in+Interactive+Evolutionary+Multiobjective+Optimization&rft.jtitle=IEEE+transactions+on+evolutionary+computation&rft.au=Branke%2C+Jurgen&rft.au=Greco%2C+Salvatore&rft.au=Slowinski%2C+Roman&rft.au=Zielniewicz%2C+Piotr&rft.date=2015-02-01&rft.issn=1089-778X&rft.eissn=1941-0026&rft.volume=19&rft.issue=1&rft.spage=88&rft.epage=102&rft_id=info:doi/10.1109%2FTEVC.2014.2303783&rft.externalDBID=n%2Fa&rft.externalDocID=10_1109_TEVC_2014_2303783
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