Convergence analysis of stochastic higher-order majorization-minimization algorithms

Majorization-minimization schemes are a broad class of iterative methods targeting general optimization problems, including nonconvex, nonsmooth and stochastic. These algorithms minimize successively a sequence of upper bounds of the objective function so that along the iterations the objective valu...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Optimization methods & software Ročník 39; číslo 2; s. 384 - 413
Hlavní autoři: Lupu, Daniela, Necoara, Ion
Médium: Journal Article
Jazyk:angličtina
Vydáno: Abingdon Taylor & Francis 03.03.2024
Taylor & Francis Ltd
Témata:
ISSN:1055-6788, 1029-4937
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Abstract Majorization-minimization schemes are a broad class of iterative methods targeting general optimization problems, including nonconvex, nonsmooth and stochastic. These algorithms minimize successively a sequence of upper bounds of the objective function so that along the iterations the objective value decreases. We present a stochastic higher-order algorithmic framework for minimizing the average of a very large number of sufficiently smooth functions. Our stochastic framework is based on the notion of stochastic higher-order upper bound approximations of the finite-sum objective function and minibatching. We derive convergence results for nonconvex and convex optimization problems when the higher-order approximation of the objective function yields an error that is p times differentiable and has Lipschitz continuous p derivative. More precisely, for general nonconvex problems we present asymptotic stationary point guarantees and under Kurdyka-Lojasiewicz property we derive local convergence rates ranging from sublinear to linear. For convex problems with uniformly convex objective function, we derive local (super)linear convergence results for our algorithm. Compared to existing stochastic (first-order) methods, our algorithm adapts to the problem's curvature and allows using any batch size. Preliminary numerical tests support the effectiveness of our algorithmic framework.
AbstractList Majorization–minimization schemes are a broad class of iterative methods targeting general optimization problems, including nonconvex, nonsmooth and stochastic. These algorithms minimize successively a sequence of upper bounds of the objective function so that along the iterations the objective value decreases. We present a stochastic higher-order algorithmic framework for minimizing the average of a very large number of sufficiently smooth functions. Our stochastic framework is based on the notion of stochastic higher-order upper bound approximations of the finite-sum objective function and minibatching. We derive convergence results for nonconvex and convex optimization problems when the higher-order approximation of the objective function yields an error that is p times differentiable and has Lipschitz continuous p derivative. More precisely, for general nonconvex problems we present asymptotic stationary point guarantees and under Kurdyka–Lojasiewicz property we derive local convergence rates ranging from sublinear to linear. For convex problems with uniformly convex objective function, we derive local (super)linear convergence results for our algorithm. Compared to existing stochastic (first-order) methods, our algorithm adapts to the problem's curvature and allows using any batch size. Preliminary numerical tests support the effectiveness of our algorithmic framework.
Author Necoara, Ion
Lupu, Daniela
Author_xml – sequence: 1
  givenname: Daniela
  surname: Lupu
  fullname: Lupu, Daniela
  organization: University Politehnica Bucharest
– sequence: 2
  givenname: Ion
  surname: Necoara
  fullname: Necoara, Ion
  email: ion.necoara@upb.ro
  organization: Gheorghe Mihoc-Caius Iacob Institute of Mathematical Statistics and Applied Mathematics of the Romanian Academy
BookMark eNqFkM1OwzAQhC1UJNrCIyBF4pyyjmMnFRdQxZ9UiUs5W45jN64Su9guqDw9CS0XDnDaXe3MavaboJF1ViF0iWGGoYRrDJSyoixnGWRklmWU5XlxgsYYsnmaz0kxGnpK00F0hiYhbAAgxzkbo9XC2Xfl18pKlQgr2n0wIXE6CdHJRoRoZNKYdaN86nytfNKJjfPmU0TjbNoZa7rjkIh23W9i04VzdKpFG9TFsU7R68P9avGULl8enxd3y1QSUsZUYVlLIpioCpIVWNEMcCmhlgKgpgoTzWrQumalLBhVpMo0JlU9V2yeMwoVmaKrw92td287FSLfuJ3vnwicQEF6Cjkpe9XNQSW9C8ErzaWJ35GjF6blGPiAkf9g5ANGfsTYu-kv99abTvj9v77bg89Y7XwnPpxvax7FvnVee2Gl6UP-feILl4SNTA
CitedBy_id crossref_primary_10_1109_TCI_2022_3230584
crossref_primary_10_1109_TCI_2025_3555137
Cites_doi 10.1561/2200000016
10.1137/140957639
10.1137/070704277
10.1137/16M1080173
10.1007/s10107-012-0629-5
10.1137/17M1122943
10.1080/10556788.2019.1678033
10.1145/1961189.1961199
10.1137/050644641
10.1093/imanum/dry009
10.1007/s10107-006-0706-8
10.1007/s10107-016-1065-8
10.1007/s10107-019-01449-1
10.1007/s00245-019-09617-7
10.1007/s10957-021-01821-2
10.1007/s10957-020-01653-6
10.1007/s10107-020-01606-x
10.1016/j.jprocont.2010.12.010
10.1080/10556788.2020.1854252
10.1002/0471221317
10.1007/s10107-004-0559-y
ContentType Journal Article
Copyright 2023 Informa UK Limited, trading as Taylor & Francis Group 2023
2023 Informa UK Limited, trading as Taylor & Francis Group
Copyright_xml – notice: 2023 Informa UK Limited, trading as Taylor & Francis Group 2023
– notice: 2023 Informa UK Limited, trading as Taylor & Francis Group
DBID AAYXX
CITATION
7SC
8FD
JQ2
L7M
L~C
L~D
DOI 10.1080/10556788.2023.2256447
DatabaseName CrossRef
Computer and Information Systems Abstracts
Technology Research Database
ProQuest Computer Science Collection
Advanced Technologies Database with Aerospace
Computer and Information Systems Abstracts – Academic
Computer and Information Systems Abstracts Professional
DatabaseTitle CrossRef
Computer and Information Systems Abstracts
Technology Research Database
Computer and Information Systems Abstracts – Academic
Advanced Technologies Database with Aerospace
ProQuest Computer Science Collection
Computer and Information Systems Abstracts Professional
DatabaseTitleList Computer and Information Systems Abstracts

DeliveryMethod fulltext_linktorsrc
Discipline Engineering
EISSN 1029-4937
EndPage 413
ExternalDocumentID 10_1080_10556788_2023_2256447
2256447
Genre Research Article
GrantInformation_xml – fundername: UEFISCDI
  grantid: NO Grants 2014-2021 RO-NO-2019-0184, under project ELO-Hyp, contract no. 24/2020
GroupedDBID .4S
.7F
.DC
.QJ
0BK
0R~
123
29N
30N
4.4
AAENE
AAGDL
AAHIA
AAJMT
AALDU
AAMIU
AAPUL
AAQRR
ABCCY
ABDBF
ABFIM
ABHAV
ABJNI
ABLIJ
ABPAQ
ABPEM
ABTAI
ABXUL
ABXYU
ACGEJ
ACGFS
ACIWK
ACTIO
ACUHS
ADCVX
ADGTB
ADXPE
AEISY
AENEX
AEOZL
AEPSL
AEYOC
AFKVX
AFRVT
AGDLA
AGMYJ
AHDZW
AIJEM
AIYEW
AJWEG
AKBVH
AKOOK
ALMA_UNASSIGNED_HOLDINGS
ALQZU
AMVHM
AQRUH
AQTUD
ARCSS
AVBZW
AWYRJ
BLEHA
CCCUG
CE4
CS3
DGEBU
DKSSO
DU5
EAP
EBS
EDO
EMK
EPL
EST
ESX
E~A
E~B
F5P
GTTXZ
H13
HF~
HZ~
H~P
I-F
IPNFZ
J.P
KYCEM
LJTGL
M4Z
NA5
NY~
O9-
P2P
PQQKQ
RIG
RNANH
ROSJB
RTWRZ
S-T
SNACF
TASJS
TBQAZ
TDBHL
TEJ
TFL
TFT
TFW
TTHFI
TUROJ
TUS
TWF
UT5
UU3
ZGOLN
~S~
AAYXX
CITATION
7SC
8FD
JQ2
L7M
L~C
L~D
ID FETCH-LOGICAL-c338t-e1cdc3a6ab73271e52018c0dca00d5e13f6d0ffd68c765e3b2f13bd9e694650b3
IEDL.DBID TFW
ISICitedReferencesCount 1
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001119296800001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 1055-6788
IngestDate Wed Aug 13 07:39:13 EDT 2025
Tue Nov 18 22:02:54 EST 2025
Sat Nov 29 02:36:07 EST 2025
Mon Oct 20 23:48:06 EDT 2025
IsPeerReviewed true
IsScholarly true
Issue 2
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-c338t-e1cdc3a6ab73271e52018c0dca00d5e13f6d0ffd68c765e3b2f13bd9e694650b3
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
PQID 3073447438
PQPubID 186278
PageCount 30
ParticipantIDs proquest_journals_3073447438
crossref_primary_10_1080_10556788_2023_2256447
crossref_citationtrail_10_1080_10556788_2023_2256447
informaworld_taylorfrancis_310_1080_10556788_2023_2256447
PublicationCentury 2000
PublicationDate 2024-03-03
PublicationDateYYYYMMDD 2024-03-03
PublicationDate_xml – month: 03
  year: 2024
  text: 2024-03-03
  day: 03
PublicationDecade 2020
PublicationPlace Abingdon
PublicationPlace_xml – name: Abingdon
PublicationTitle Optimization methods & software
PublicationYear 2024
Publisher Taylor & Francis
Taylor & Francis Ltd
Publisher_xml – name: Taylor & Francis
– name: Taylor & Francis Ltd
References e_1_3_3_30_1
e_1_3_3_18_1
e_1_3_3_17_1
e_1_3_3_19_1
e_1_3_3_13_1
e_1_3_3_16_1
e_1_3_3_35_1
e_1_3_3_15_1
e_1_3_3_36_1
e_1_3_3_34_1
e_1_3_3_12_1
e_1_3_3_31_1
e_1_3_3_11_1
e_1_3_3_32_1
Tripuraneni N. (e_1_3_3_33_1) 2018; 31
Defazio A. (e_1_3_3_10_1) 2014; 27
Moulines E. (e_1_3_3_20_1) 2011; 24
e_1_3_3_7_1
e_1_3_3_6_1
e_1_3_3_9_1
e_1_3_3_8_1
e_1_3_3_29_1
e_1_3_3_28_1
Goodfellow I. (e_1_3_3_14_1) 2016
e_1_3_3_25_1
e_1_3_3_24_1
e_1_3_3_27_1
e_1_3_3_26_1
e_1_3_3_3_1
e_1_3_3_21_1
e_1_3_3_2_1
e_1_3_3_5_1
e_1_3_3_23_1
e_1_3_3_4_1
e_1_3_3_22_1
References_xml – ident: e_1_3_3_2_1
– volume: 27
  start-page: 1646
  year: 2014
  ident: e_1_3_3_10_1
  article-title: SAGA: a fast incremental gradient method with support for non-strongly convex composite objectives
  publication-title: Adv. Neural. Inf. Process. Syst.
– ident: e_1_3_3_30_1
– ident: e_1_3_3_7_1
  doi: 10.1561/2200000016
– volume: 24
  start-page: 451
  year: 2011
  ident: e_1_3_3_20_1
  article-title: Non-asymptotic analysis of stochastic approximation algorithms for machine learning
  publication-title: Adv. Neural Inform. Process. Syst.
– ident: e_1_3_3_21_1
  doi: 10.1137/140957639
– ident: e_1_3_3_28_1
  doi: 10.1137/070704277
– volume: 31
  start-page: 2899
  year: 2018
  ident: e_1_3_3_33_1
  article-title: Stochastic cubic regularization for fast nonconvex optimization
  publication-title: Adv. Neural. Inf. Process. Syst.
– ident: e_1_3_3_3_1
  doi: 10.1137/16M1080173
– ident: e_1_3_3_24_1
  doi: 10.1007/s10107-012-0629-5
– ident: e_1_3_3_19_1
  doi: 10.1137/17M1122943
– ident: e_1_3_3_8_1
  doi: 10.1080/10556788.2019.1678033
– ident: e_1_3_3_13_1
– ident: e_1_3_3_9_1
  doi: 10.1145/1961189.1961199
– ident: e_1_3_3_16_1
– ident: e_1_3_3_6_1
  doi: 10.1137/050644641
– ident: e_1_3_3_5_1
  doi: 10.1093/imanum/dry009
– ident: e_1_3_3_35_1
– ident: e_1_3_3_27_1
– ident: e_1_3_3_29_1
  doi: 10.1007/s10107-006-0706-8
– ident: e_1_3_3_4_1
  doi: 10.1007/s10107-016-1065-8
– ident: e_1_3_3_18_1
– ident: e_1_3_3_12_1
– ident: e_1_3_3_22_1
  doi: 10.1007/s10107-019-01449-1
– ident: e_1_3_3_32_1
  doi: 10.1007/s00245-019-09617-7
– ident: e_1_3_3_25_1
  doi: 10.1007/s10957-021-01821-2
– ident: e_1_3_3_31_1
  doi: 10.1007/s10957-020-01653-6
– ident: e_1_3_3_17_1
– ident: e_1_3_3_11_1
  doi: 10.1007/s10107-020-01606-x
– ident: e_1_3_3_34_1
– ident: e_1_3_3_26_1
  doi: 10.1016/j.jprocont.2010.12.010
– ident: e_1_3_3_23_1
  doi: 10.1080/10556788.2020.1854252
– volume-title: Deep Learning
  year: 2016
  ident: e_1_3_3_14_1
– ident: e_1_3_3_15_1
  doi: 10.1002/0471221317
– ident: e_1_3_3_36_1
  doi: 10.1007/s10107-004-0559-y
SSID ssj0004146
Score 2.3532667
Snippet Majorization-minimization schemes are a broad class of iterative methods targeting general optimization problems, including nonconvex, nonsmooth and...
Majorization–minimization schemes are a broad class of iterative methods targeting general optimization problems, including nonconvex, nonsmooth and...
SourceID proquest
crossref
informaworld
SourceType Aggregation Database
Enrichment Source
Index Database
Publisher
StartPage 384
SubjectTerms Algorithms
Approximation
Convergence
convergence rates
Convexity
Finite-sum optimization
Iterative methods
majorization-minimization
minibatch
Optimization
stochastic higher-order algorithms
Upper bounds
Title Convergence analysis of stochastic higher-order majorization-minimization algorithms
URI https://www.tandfonline.com/doi/abs/10.1080/10556788.2023.2256447
https://www.proquest.com/docview/3073447438
Volume 39
WOSCitedRecordID wos001119296800001&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: PRVAWR
  databaseName: Taylor & Francis Online Journals
  customDbUrl:
  eissn: 1029-4937
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0004146
  issn: 1055-6788
  databaseCode: TFW
  dateStart: 19920101
  isFulltext: true
  titleUrlDefault: https://www.tandfonline.com
  providerName: Taylor & Francis
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwrV07T8MwELZQxQADb0ShIA-sLkmdOPGIKiqmiqGIbpGfFNQHagK_n7PjQCuEOsAQRZZ1VnRn--6i775D6DrvcZtaxogSkSZJljPCKYsIRNqZlFwoeHyziWw4zMdj_hDQhGWAVboc2tZEEf6udodbyLJBxN34no6QunVd6-8ubEjw6a6eHCJ7B-obDZ6-KyNDfRFIECfS1PD8tsqad1rjLv1xV3sHNNj_h08_QHsh-sS39XY5RFtmfoR2VzgJj9Go73DoviTTYBEoS_DCYogS1UQ4Wmc88eAQ4mk78Uy8LpahmpM4ppJZGGAxfYaZajIrT9Dj4G7Uvyeh9QJRkLNWxMRKKyqYkBntZbFJIU7IVaTBopFOTUwt05G1muUqY6mhsmdjKjU3jCcQ80l6ilrzxdycIZwqCV4Q8iZpokTlisObCSM1U0zELGqjpFF5oQIvuWuPMS3iQF_aKK1wSiuC0tqo-yX2VhNzbBLgq_YsKv9HxNbtSwq6QbbTGL8IZxxE4HaEuYTm539Y-gLtwDDxsDbaQa1q-W4u0bb6qF7K5ZXfzZ-t5vEi
linkProvider Taylor & Francis
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwrV1LT8MwDI5gIAEH3ojBgB64ZrRLm7ZHhJiGGDsVsVuUpAkD7YG2wu_HSVPYhNAOcKiqynJUOQ_bkf19CF0mrVRHmlIsuZ_jME4oTgn1MUTasRApl_BYsom410v6_XS-F8aUVZocWpdAEfasNpvbXEZXJXFXltQRcrem4f5uwooEpx6vojXDTmcSsKz99N0b6TqMQAUbnaqL57dhFvzTAnrpj9PauqD2zn_8_C7adgGod12umD20osb7aGsOlvAAZTemFN12ZSqPO9QSb6I9CBTlgBtkZ29g60OwRe70Rvx1MnUNndiAlYzch8eHzyApBqPZIXps32Y3HezYF7CEtLXAKpC5JJxyEZNWHKgIQoVE-jlMqp9HKiCa5r7WOU1kTCNFREsHROSpomkIYZ8gR6g2nozVMfIiKcARQuoklB_KRKbwplyJnErKA-rXUVjZnEkHTW4YMoYscAimldGYMRpzRquj5pfaW4nNsUwhnZ9QVthLEV0ymDCyRLdRzT5z2xxU4IAEWUiSkz8MfYE2OtlDl3XvevenaBNEoa1yIw1UK6bv6gyty4_iZTY9t0v7E-XY9UU
linkToPdf http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwrV1LT8MwDI5gIAQH3ojBgB64ZrRLm7ZHNJhAoGmHIXar8mSgvbQVfj9OmsImhHaAQ1VVlqPKdmI7sj8jdJk0Uh1pSrFgvsRhnFCcEupjiLRjzlMm4LHDJuJ2O-n10o6rJpy5skqTQ-sCKMKe1WZzT6QuK-Ku7ExHSN3qZvR3HQwSfHq8itYgdI6MYXdbz9-tka7BCFiw4SmbeH5bZsE9LYCX_jisrQdq7fzDv--ibRd-eteFveyhFTXaR1tzoIQHqNs0hei2J1N5zGGWeGPtQZgo-szgOnt9Wx2CLW6nN2Rv46lr58QGqmToPjw2eAFK3h_ODtFT67bbvMNu9gIWkLTmWAVCCsIo4zFpxIGKIFBIhC9Bpb6MVEA0lb7WkiYCpK4Ib-iAcJkqmoYQ9HFyhCqj8UgdIy8SHNwgJE5c-aFIRApvyhSXVFAWUL-KwlLkmXDA5GY-xiALHH5pKbTMCC1zQqui-hfbpEDmWMaQzuszy-2ViC7ml2RkCW-tVH7mNjmwwPEItJAkJ39Y-gJtdG5a2eN9--EUbQIltCVupIYq-fRdnaF18ZG_zqbn1rA_AQco8_c
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=Convergence+analysis+of+stochastic+higher-order+majorization-minimization+algorithms&rft.jtitle=Optimization+methods+%26+software&rft.au=Lupu%2C+Daniela&rft.au=Necoara%2C+Ion&rft.date=2024-03-03&rft.pub=Taylor+%26+Francis&rft.issn=1055-6788&rft.eissn=1029-4937&rft.volume=39&rft.issue=2&rft.spage=384&rft.epage=413&rft_id=info:doi/10.1080%2F10556788.2023.2256447&rft.externalDocID=2256447
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1055-6788&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1055-6788&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1055-6788&client=summon