Learning-augmented maximum flow

We propose a framework for speeding up maximum flow computation by using predictions. A prediction is a flow, i.e., an assignment of non-negative flow values to edges, which satisfies the flow conservation property, but does not necessarily respect the edge capacities of the actual instance (since t...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Information processing letters Jg. 186; S. 106487
Hauptverfasser: Polak, Adam, Zub, Maksym
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Elsevier B.V 01.08.2024
Schlagworte:
ISSN:0020-0190, 1872-6119
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Abstract We propose a framework for speeding up maximum flow computation by using predictions. A prediction is a flow, i.e., an assignment of non-negative flow values to edges, which satisfies the flow conservation property, but does not necessarily respect the edge capacities of the actual instance (since these were unknown at the time of learning). We present an algorithm that, given an m-edge flow network and a predicted flow, computes a maximum flow in O(mη) time, where η is the ℓ1 error of the prediction, i.e., the sum over the edges of the absolute difference between the predicted and optimal flow values. Moreover, we prove that, given an oracle access to a distribution over flow networks, it is possible to efficiently PAC-learn a prediction minimizing the expected ℓ1 error over that distribution. Our results fit into the recent line of research on learning-augmented algorithms, which aims to improve over worst-case bounds of classical algorithms by using predictions, e.g., machine-learned from previous similar instances. So far, the main focus in this area was on improving competitive ratios for online problems. Following Dinitz et al. (2021) [6], our results are among the firsts to improve the running time of an offline problem. •We speed up maximum flow computation in graphs by using predictions.•Prediction is a flow satisfying flow conservation but not necessarily capacities.•We show that such predictions are efficiently learnable.•We compute flow in O(mη) time in m-edge graphs for predictions with L1 error up to η.
AbstractList We propose a framework for speeding up maximum flow computation by using predictions. A prediction is a flow, i.e., an assignment of non-negative flow values to edges, which satisfies the flow conservation property, but does not necessarily respect the edge capacities of the actual instance (since these were unknown at the time of learning). We present an algorithm that, given an m-edge flow network and a predicted flow, computes a maximum flow in O(mη) time, where η is the ℓ1 error of the prediction, i.e., the sum over the edges of the absolute difference between the predicted and optimal flow values. Moreover, we prove that, given an oracle access to a distribution over flow networks, it is possible to efficiently PAC-learn a prediction minimizing the expected ℓ1 error over that distribution. Our results fit into the recent line of research on learning-augmented algorithms, which aims to improve over worst-case bounds of classical algorithms by using predictions, e.g., machine-learned from previous similar instances. So far, the main focus in this area was on improving competitive ratios for online problems. Following Dinitz et al. (2021) [6], our results are among the firsts to improve the running time of an offline problem. •We speed up maximum flow computation in graphs by using predictions.•Prediction is a flow satisfying flow conservation but not necessarily capacities.•We show that such predictions are efficiently learnable.•We compute flow in O(mη) time in m-edge graphs for predictions with L1 error up to η.
ArticleNumber 106487
Author Polak, Adam
Zub, Maksym
Author_xml – sequence: 1
  givenname: Adam
  surname: Polak
  fullname: Polak, Adam
  email: adam.polak@unibocconi.it
  organization: Bocconi University, Via Roentgen 1, 20136, Milano, Italy
– sequence: 2
  givenname: Maksym
  surname: Zub
  fullname: Zub, Maksym
  email: max.zub@student.uj.edu.pl
  organization: Jagiellonian University, ul. Lojasiewicza 6, 30-348, Krakow, Poland
BookMark eNp9j8tOwzAQRS1UJNrCB7CiP5AwkzhxLFaogoJUiU33lmNPKkeJUzkpj7_HVVmx6GruaOZc6SzYzA-eGLtHSBGwfGxTd-jSDDIe95JX4orNsRJZUiLKGZsDZJAASrhhi3FsAeJTLubsYUs6eOf3iT7ue_IT2VWvv11_7FdNN3zdsutGdyPd_c0l272-7NZvyfZj875-3iYm5zAlhNaiLhowkpMpaxlDmde5yCpeWSmEbEjWGTfSWm4LizwvKF401lzyIl8yPNeaMIxjoEYdgut1-FEI6iSoWhUF1UlQnQUjI_4xxk16coOfgnbdRfLpTFI0-nQU1GgceUPWBTKTsoO7QP8CLdRr2Q
CitedBy_id crossref_primary_10_3390_s25082555
Cites_doi 10.4153/CJM-1956-045-5
10.1109/TPAMI.2004.60
10.1145/3447579
10.1016/0022-0000(80)90035-5
10.1007/s11554-013-0344-3
10.1145/321694.321699
10.1080/01621459.1963.10500830
10.1287/opre.1080.0524
ContentType Journal Article
Copyright 2024 Elsevier B.V.
Copyright_xml – notice: 2024 Elsevier B.V.
DBID AAYXX
CITATION
DOI 10.1016/j.ipl.2024.106487
DatabaseName CrossRef
DatabaseTitle CrossRef
DatabaseTitleList
DeliveryMethod fulltext_linktorsrc
Discipline Computer Science
EISSN 1872-6119
ExternalDocumentID 10_1016_j_ipl_2024_106487
S0020019024000176
GroupedDBID --K
--M
-~X
.DC
.~1
0R~
1B1
1RT
1~.
1~5
29I
4.4
457
4G.
5GY
5VS
7-5
71M
8P~
9JN
AACTN
AAEDT
AAEDW
AAIAV
AAIKJ
AAKOC
AALRI
AAOAW
AAQFI
AAQXK
AAXUO
AAYFN
ABBOA
ABEFU
ABFNM
ABFSI
ABJNI
ABMAC
ABTAH
ABXDB
ACDAQ
ACGFS
ACNNM
ACRLP
ACZNC
ADBBV
ADEZE
ADJOM
ADMUD
AEBSH
AEKER
AENEX
AFKWA
AFTJW
AGHFR
AGUBO
AGYEJ
AHHHB
AHZHX
AIALX
AIEXJ
AIKHN
AITUG
AJOXV
AKRWK
ALMA_UNASSIGNED_HOLDINGS
AMFUW
AMRAJ
AOUOD
ASPBG
AVWKF
AXJTR
AZFZN
BKOJK
BKOMP
BLXMC
CS3
DU5
E.L
EBS
EFJIC
EJD
EO8
EO9
EP2
EP3
F5P
FDB
FEDTE
FGOYB
FIRID
FNPLU
FYGXN
G-2
G-Q
G8K
GBLVA
GBOLZ
HLZ
HMJ
HVGLF
HZ~
IHE
J1W
KOM
LG9
M26
M41
MO0
MS~
O-L
O9-
OAUVE
OZT
P-8
P-9
P2P
PC.
PQQKQ
Q38
R2-
RIG
ROL
RPZ
SBC
SDF
SDG
SDP
SES
SEW
SME
SPC
SPCBC
SSV
SSZ
T5K
TN5
UQL
WH7
WUQ
XPP
ZMT
ZY4
~G-
9DU
AATTM
AAXKI
AAYWO
AAYXX
ABDPE
ABWVN
ACLOT
ACRPL
ACVFH
ADCNI
ADNMO
AEIPS
AEUPX
AFJKZ
AFPUW
AGQPQ
AIGII
AIIUN
AKBMS
AKYEP
ANKPU
APXCP
CITATION
EFKBS
EFLBG
~HD
ID FETCH-LOGICAL-c340t-e1dd1a5f0c94ec6b90c963b372848d9779fe9b24c9dd4d5d1435e48da1b49453
ISICitedReferencesCount 2
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001253808800001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 0020-0190
IngestDate Sat Nov 29 03:44:24 EST 2025
Tue Nov 18 22:19:47 EST 2025
Sat May 11 15:33:44 EDT 2024
IsDoiOpenAccess false
IsOpenAccess true
IsPeerReviewed true
IsScholarly true
Keywords Combinatorial optimization
Analysis of algorithms
Algorithms with predictions
Maximum flow
Language English
LinkModel OpenURL
MergedId FETCHMERGED-LOGICAL-c340t-e1dd1a5f0c94ec6b90c963b372848d9779fe9b24c9dd4d5d1435e48da1b49453
OpenAccessLink https://ruj.uj.edu.pl/xmlui/handle/item/327649
ParticipantIDs crossref_primary_10_1016_j_ipl_2024_106487
crossref_citationtrail_10_1016_j_ipl_2024_106487
elsevier_sciencedirect_doi_10_1016_j_ipl_2024_106487
PublicationCentury 2000
PublicationDate August 2024
2024-08-00
PublicationDateYYYYMMDD 2024-08-01
PublicationDate_xml – month: 08
  year: 2024
  text: August 2024
PublicationDecade 2020
PublicationTitle Information processing letters
PublicationYear 2024
Publisher Elsevier B.V
Publisher_xml – name: Elsevier B.V
References Chen, Silwal, Vakilian, Zhang (br0030) 2022; vol. 162
Spielman, Teng (br0260) 2004
Kohli, Torr (br0190) 2005
Hoeffding (br0160) 1963; 58
Edmonds, Karp (br0090) 1972; 19
Lykouris, Vassilvitskii (br0210) 2021; 68
Purohit, Svitkina, Kumar (br0240) 2018
Hsu, Indyk, Katabi, Vakilian (br0170) 2019
Ford, Fulkerson (br0120) 1956; 8
Dinitz (br0070) 1970; 11
Juan, Boykov (br0180) 2006
Boykov, Kolmogorov (br0020) 2004; 26
Galil, Naamad (br0130) 1980; 21
Hochbaum (br0150) 2008; 56
Chen, Kyng, Liu, Peng, Probst Gutenberg, Sachdeva (br0040) 2022
Dinitz, Im, Lavastida, Moseley, Vassilvitskii (br0060) 2021; vol. 34
Ergun, Feng, Silwal, Woodruff, Zhou (br0100) 2022
Kraska, Beutel, Chi, Dean, Polyzotis (br0200) 2018
Fishbain, Hochbaum, Müller (br0110) 2016; 11
Goldberg, Tarjan (br0140) 1986
Sakaue, Oki (br0250) 2022
Eden, Indyk, Narayanan, Rubinfeld, Silwal, Wagner (br0080) 2021
Mitzenmacher, Vassilvitskii (br0230) 2020
Davies, Moseley, Vassilvitskii, Wang (br0050) 2023; vol. 202
Ahuja, Magnanti, Orlin (br0010) 1993
Mahdian, Nazerzadeh, Saberi (br0220) 2007
Ergun (10.1016/j.ipl.2024.106487_br0100) 2022
Sakaue (10.1016/j.ipl.2024.106487_br0250) 2022
Dinitz (10.1016/j.ipl.2024.106487_br0060) 2021; vol. 34
Boykov (10.1016/j.ipl.2024.106487_br0020) 2004; 26
Eden (10.1016/j.ipl.2024.106487_br0080) 2021
Hoeffding (10.1016/j.ipl.2024.106487_br0160) 1963; 58
Galil (10.1016/j.ipl.2024.106487_br0130) 1980; 21
Mitzenmacher (10.1016/j.ipl.2024.106487_br0230) 2020
Dinitz (10.1016/j.ipl.2024.106487_br0070) 1970; 11
Purohit (10.1016/j.ipl.2024.106487_br0240) 2018
Fishbain (10.1016/j.ipl.2024.106487_br0110) 2016; 11
Chen (10.1016/j.ipl.2024.106487_br0030) 2022; vol. 162
Edmonds (10.1016/j.ipl.2024.106487_br0090) 1972; 19
Kohli (10.1016/j.ipl.2024.106487_br0190) 2005
Hsu (10.1016/j.ipl.2024.106487_br0170) 2019
Juan (10.1016/j.ipl.2024.106487_br0180) 2006
Mahdian (10.1016/j.ipl.2024.106487_br0220) 2007
Davies (10.1016/j.ipl.2024.106487_br0050) 2023; vol. 202
Kraska (10.1016/j.ipl.2024.106487_br0200) 2018
Spielman (10.1016/j.ipl.2024.106487_br0260) 2004
Lykouris (10.1016/j.ipl.2024.106487_br0210) 2021; 68
Ford (10.1016/j.ipl.2024.106487_br0120) 1956; 8
Chen (10.1016/j.ipl.2024.106487_br0040) 2022
Ahuja (10.1016/j.ipl.2024.106487_br0010) 1993
Goldberg (10.1016/j.ipl.2024.106487_br0140) 1986
Hochbaum (10.1016/j.ipl.2024.106487_br0150) 2008; 56
References_xml – start-page: 612
  year: 2022
  end-page: 623
  ident: br0040
  article-title: Maximum flow and minimum-cost flow in almost-linear time
  publication-title: 63rd IEEE Annual Symposium on Foundations of Computer Science
– volume: 58
  start-page: 13
  year: 1963
  end-page: 30
  ident: br0160
  article-title: Probability inequalities for sums of bounded random variables
  publication-title: J. Am. Stat. Assoc.
– year: 1993
  ident: br0010
  article-title: Network Flows – Theory, Algorithms and Applications
– volume: vol. 34
  start-page: 10393
  year: 2021
  end-page: 10406
  ident: br0060
  article-title: Faster Matchings via Learned Duals
  publication-title: Advances in Neural Information Processing Systems
– volume: 11
  start-page: 1277
  year: 1970
  end-page: 1280
  ident: br0070
  article-title: Algorithm for solution of a problem of maximum flow in networks with power estimation
  publication-title: Sov. Math. Dokl.
– year: 2022
  ident: br0100
  article-title: Learning-augmented k-means clustering
  publication-title: The Tenth International Conference on Learning Representations
– volume: vol. 162
  start-page: 3583
  year: 2022
  end-page: 3602
  ident: br0030
  article-title: Faster fundamental graph algorithms via learned predictions
  publication-title: International Conference on Machine Learning
– start-page: 9684
  year: 2018
  end-page: 9693
  ident: br0240
  article-title: Improving online algorithms via ML predictions
  publication-title: Advances in Neural Information Processing Systems 31: Annual Conference on Neural Information Processing Systems 2018
– year: 2019
  ident: br0170
  article-title: Learning-based frequency estimation algorithms
  publication-title: 7th International Conference on Learning Representations
– start-page: 489
  year: 2018
  end-page: 504
  ident: br0200
  article-title: The case for learned index structures
  publication-title: Proceedings of the 2018 International Conference on Management of Data
– volume: 26
  start-page: 1124
  year: 2004
  end-page: 1137
  ident: br0020
  article-title: An experimental comparison of min-cut/max-flow algorithms for energy minimization in vision
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
– volume: 56
  start-page: 992
  year: 2008
  end-page: 1009
  ident: br0150
  article-title: The pseudoflow algorithm: a new algorithm for the maximum-flow problem
  publication-title: Oper. Res.
– volume: 11
  start-page: 589
  year: 2016
  end-page: 609
  ident: br0110
  article-title: A competitive study of the pseudoflow algorithm for the minimum s-t cut problem in vision applications
  publication-title: J. Real-Time Image Process.
– volume: 19
  start-page: 248
  year: 1972
  end-page: 264
  ident: br0090
  article-title: Theoretical improvements in algorithmic efficiency for network flow problems
  publication-title: J. ACM
– start-page: 136
  year: 1986
  end-page: 146
  ident: br0140
  article-title: A new approach to the maximum flow problem
  publication-title: Proceedings of the 18th Annual ACM Symposium on Theory of Computing
– start-page: 922
  year: 2005
  end-page: 929
  ident: br0190
  article-title: Efficiently solving dynamic Markov random fields using graph cuts
  publication-title: 10th IEEE International Conference on Computer Vision (ICCV 2005)
– volume: 68
  year: 2021
  ident: br0210
  article-title: Competitive caching with machine learned advice
  publication-title: J. ACM
– volume: vol. 202
  start-page: 7231
  year: 2023
  end-page: 7248
  ident: br0050
  article-title: Predictive flows for faster Ford-Fulkerson
  publication-title: International Conference on Machine Learning
– volume: 8
  start-page: 399
  year: 1956
  end-page: 404
  ident: br0120
  article-title: Maximal flow through a network
  publication-title: Can. J. Math.
– start-page: 81
  year: 2004
  end-page: 90
  ident: br0260
  article-title: Nearly-linear time algorithms for graph partitioning, graph sparsification, and solving linear systems
  publication-title: Proceedings of the 36th Annual ACM Symposium on Theory of Computing
– year: 2021
  ident: br0080
  article-title: Learning-based support estimation in sublinear time
  publication-title: 9th International Conference on Learning Representations
– volume: 21
  start-page: 203
  year: 1980
  end-page: 217
  ident: br0130
  article-title: An
  publication-title: J. Comput. Syst. Sci.
– year: 2022
  ident: br0250
  article-title: Discrete-convex-analysis-based framework for warm-starting algorithms with predictions
  publication-title: NeurIPS
– start-page: 646
  year: 2020
  end-page: 662
  ident: br0230
  article-title: Algorithms with predictions
  publication-title: Beyond the Worst-Case Analysis of Algorithms
– start-page: 1023
  year: 2006
  end-page: 1029
  ident: br0180
  article-title: Active graph cuts
  publication-title: 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR 2006)
– start-page: 288
  year: 2007
  end-page: 294
  ident: br0220
  article-title: Allocating online advertisement space with unreliable estimates
  publication-title: Proceedings 8th ACM Conference on Electronic Commerce (EC-2007)
– year: 2022
  ident: 10.1016/j.ipl.2024.106487_br0250
  article-title: Discrete-convex-analysis-based framework for warm-starting algorithms with predictions
– volume: 8
  start-page: 399
  year: 1956
  ident: 10.1016/j.ipl.2024.106487_br0120
  article-title: Maximal flow through a network
  publication-title: Can. J. Math.
  doi: 10.4153/CJM-1956-045-5
– start-page: 288
  year: 2007
  ident: 10.1016/j.ipl.2024.106487_br0220
  article-title: Allocating online advertisement space with unreliable estimates
– year: 1993
  ident: 10.1016/j.ipl.2024.106487_br0010
– start-page: 9684
  year: 2018
  ident: 10.1016/j.ipl.2024.106487_br0240
  article-title: Improving online algorithms via ML predictions
– volume: 26
  start-page: 1124
  issue: 9
  year: 2004
  ident: 10.1016/j.ipl.2024.106487_br0020
  article-title: An experimental comparison of min-cut/max-flow algorithms for energy minimization in vision
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
  doi: 10.1109/TPAMI.2004.60
– year: 2022
  ident: 10.1016/j.ipl.2024.106487_br0100
  article-title: Learning-augmented k-means clustering
– volume: 68
  issue: 4
  year: 2021
  ident: 10.1016/j.ipl.2024.106487_br0210
  article-title: Competitive caching with machine learned advice
  publication-title: J. ACM
  doi: 10.1145/3447579
– start-page: 922
  year: 2005
  ident: 10.1016/j.ipl.2024.106487_br0190
  article-title: Efficiently solving dynamic Markov random fields using graph cuts
– volume: 21
  start-page: 203
  issue: 2
  year: 1980
  ident: 10.1016/j.ipl.2024.106487_br0130
  article-title: An O(EVlog2⁡V) algorithm for the maximal flow problem
  publication-title: J. Comput. Syst. Sci.
  doi: 10.1016/0022-0000(80)90035-5
– start-page: 612
  year: 2022
  ident: 10.1016/j.ipl.2024.106487_br0040
  article-title: Maximum flow and minimum-cost flow in almost-linear time
– start-page: 81
  year: 2004
  ident: 10.1016/j.ipl.2024.106487_br0260
  article-title: Nearly-linear time algorithms for graph partitioning, graph sparsification, and solving linear systems
– volume: 11
  start-page: 589
  issue: 3
  year: 2016
  ident: 10.1016/j.ipl.2024.106487_br0110
  article-title: A competitive study of the pseudoflow algorithm for the minimum s-t cut problem in vision applications
  publication-title: J. Real-Time Image Process.
  doi: 10.1007/s11554-013-0344-3
– year: 2021
  ident: 10.1016/j.ipl.2024.106487_br0080
  article-title: Learning-based support estimation in sublinear time
– start-page: 136
  year: 1986
  ident: 10.1016/j.ipl.2024.106487_br0140
  article-title: A new approach to the maximum flow problem
– volume: vol. 202
  start-page: 7231
  year: 2023
  ident: 10.1016/j.ipl.2024.106487_br0050
  article-title: Predictive flows for faster Ford-Fulkerson
– volume: vol. 34
  start-page: 10393
  year: 2021
  ident: 10.1016/j.ipl.2024.106487_br0060
  article-title: Faster Matchings via Learned Duals
– start-page: 1023
  year: 2006
  ident: 10.1016/j.ipl.2024.106487_br0180
  article-title: Active graph cuts
– volume: vol. 162
  start-page: 3583
  year: 2022
  ident: 10.1016/j.ipl.2024.106487_br0030
  article-title: Faster fundamental graph algorithms via learned predictions
– volume: 11
  start-page: 1277
  year: 1970
  ident: 10.1016/j.ipl.2024.106487_br0070
  article-title: Algorithm for solution of a problem of maximum flow in networks with power estimation
  publication-title: Sov. Math. Dokl.
– year: 2019
  ident: 10.1016/j.ipl.2024.106487_br0170
  article-title: Learning-based frequency estimation algorithms
– start-page: 489
  year: 2018
  ident: 10.1016/j.ipl.2024.106487_br0200
  article-title: The case for learned index structures
– volume: 19
  start-page: 248
  issue: 2
  year: 1972
  ident: 10.1016/j.ipl.2024.106487_br0090
  article-title: Theoretical improvements in algorithmic efficiency for network flow problems
  publication-title: J. ACM
  doi: 10.1145/321694.321699
– volume: 58
  start-page: 13
  issue: 301
  year: 1963
  ident: 10.1016/j.ipl.2024.106487_br0160
  article-title: Probability inequalities for sums of bounded random variables
  publication-title: J. Am. Stat. Assoc.
  doi: 10.1080/01621459.1963.10500830
– volume: 56
  start-page: 992
  issue: 4
  year: 2008
  ident: 10.1016/j.ipl.2024.106487_br0150
  article-title: The pseudoflow algorithm: a new algorithm for the maximum-flow problem
  publication-title: Oper. Res.
  doi: 10.1287/opre.1080.0524
– start-page: 646
  year: 2020
  ident: 10.1016/j.ipl.2024.106487_br0230
  article-title: Algorithms with predictions
SSID ssj0006437
Score 2.4305844
Snippet We propose a framework for speeding up maximum flow computation by using predictions. A prediction is a flow, i.e., an assignment of non-negative flow values...
SourceID crossref
elsevier
SourceType Enrichment Source
Index Database
Publisher
StartPage 106487
SubjectTerms Algorithms with predictions
Analysis of algorithms
Combinatorial optimization
Maximum flow
Title Learning-augmented maximum flow
URI https://dx.doi.org/10.1016/j.ipl.2024.106487
Volume 186
WOSCitedRecordID wos001253808800001&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: PRVESC
  databaseName: Elsevier SD Freedom Collection Journals 2021
  customDbUrl:
  eissn: 1872-6119
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0006437
  issn: 0020-0190
  databaseCode: AIEXJ
  dateStart: 19950113
  isFulltext: true
  titleUrlDefault: https://www.sciencedirect.com
  providerName: Elsevier
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV3LT4MwGG90evDi2_iWgycXFh4d0ONiZtQY42Exu5HSlmU62OKGzv_er7QwnM7owQshhZbC7-Prr-33QOhceNznDsOmjD1m4lhEJvWIZ9peLBzLi70o38F_vPPv74Nulzzo3HrjPJ2An6bBdEpG_wo1lAHY0nX2D3CXjUIBnAPocATY4fgr4HXE1J5Js14ecZPXEzrtJ1lSjwfDtyob1b5IuQiMlMeAXDkY5C4-Jdl-gNlvrjNbnCblQnMWKVef5_F7Ul06cHBpuDYz5ZdWaSpf50wdVhUazBixGhG_6Fo17X9q9EdyC8fBjdm9n-Naz403pRVgYWD2FEIToWwiVE0soxXHb5KghlZaN-3ubTm0yl1GZbOj-l1sU-cGe3P9-J5oVMhDZxOta9ZvtBRaW2hJpNtoo8ioYWgFu4POvoJnaPAMCd4u6ly1O5fXpk5hYTIXWxNT2JzbtBlbjGDBpOAz0HiR6wMrCDhwbxILEjmYEc4xb3LJXgVcoXaECW66e6iWDlOxjwxgpi6hjpw_BvBLYeq4TMQ2xcJmtsXEAbKK1w2ZDu8us4wMwoWf-QBdlFVGKrbJTzfj4huGmpwp0hWCPCyudviXZxyhtZmYHqPa5CUTJ2iVvU7645dTLQwfK91Wgw
linkProvider Elsevier
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-augmented+maximum+flow&rft.jtitle=Information+processing+letters&rft.au=Polak%2C+Adam&rft.au=Zub%2C+Maksym&rft.date=2024-08-01&rft.issn=0020-0190&rft.volume=186&rft.spage=106487&rft_id=info:doi/10.1016%2Fj.ipl.2024.106487&rft.externalDBID=n%2Fa&rft.externalDocID=10_1016_j_ipl_2024_106487
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0020-0190&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0020-0190&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0020-0190&client=summon