A LINEAR-PROGRAMMING APPROACH FOR THE WEIGHTED GRAPH MATCHING PROBLEM

A linear programming (LP) approach is proposed for the weighted graph matching problem. A linear program is obtained by formulating the graph matching problem in L1 norm and then transforming the resulting quadratic optimization problem to a linear one. The linear program is solved using a Simplex-b...

Full description

Saved in:
Bibliographic Details
Main Author: Al-Mohamad, HA (author)
Other Authors: Duffuaa, S. O. (author), unknown (author)
Format: article
Published: 2020
Subjects:
Online Access:https://eprints.kfupm.edu.sa/id/eprint/1985/1/a_linear_programming_approach_for_the_we_almohamad_isi_a1993lb47000012.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1864513400894128128
author Al-Mohamad, HA
author2 Duffuaa, S. O.
unknown
author2_role author
author
author_facet Al-Mohamad, HA
Duffuaa, S. O.
unknown
author_role author
dc.creator.none.fl_str_mv Al-Mohamad, HA
Duffuaa, S. O.
unknown
dc.date.*.fl_str_mv 2020
dc.format.none.fl_str_mv application/pdf
dc.identifier.none.fl_str_mv https://eprints.kfupm.edu.sa/id/eprint/1985/1/a_linear_programming_approach_for_the_we_almohamad_isi_a1993lb47000012.pdf
A LINEAR-PROGRAMMING APPROACH FOR THE WEIGHTED GRAPH MATCHING PROBLEM. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 15. pp. 522-525.
dc.language.none.fl_str_mv en
dc.publisher.none.fl_str_mv IEEE COMPUTER SOC
dc.relation.none.fl_str_mv https://eprints.kfupm.edu.sa/id/eprint/1985/
http://isi.kfupm.edu.sa/journals/pdf/A/a_linear_programming_approach_for_the_we_almohamad_isi_a1993lb47000012.pdf
dc.rights.*.fl_str_mv info:eu-repo/semantics/openAccess
dc.subject.none.fl_str_mv Systems
dc.title.none.fl_str_mv A LINEAR-PROGRAMMING APPROACH FOR THE WEIGHTED GRAPH MATCHING PROBLEM
dc.type.none.fl_str_mv Article
PeerReviewed
info:eu-repo/semantics/publishedVersion
info:eu-repo/semantics/article
description A linear programming (LP) approach is proposed for the weighted graph matching problem. A linear program is obtained by formulating the graph matching problem in L1 norm and then transforming the resulting quadratic optimization problem to a linear one. The linear program is solved using a Simplex-based algorithm. Then, approximate 0-1 integer solutions are obtained by applying the Hungarian method on the real solutions of the linear program. The complexity of the proposed algorithm is polynomial time, and it is O(n6 L) for matching graphs of size n. The developed algorithm is compared to two other algorithms. One is based on an eigen decomposition approach and the other on a symmetric polynomial transform. Experimental results showed that the LP approach is superior in matching graphs than both other methods.
eu_rights_str_mv openAccess
format article
id KFUPM_f4fcc566b24b1131c9d853604697753d
identifier_str_mv A LINEAR-PROGRAMMING APPROACH FOR THE WEIGHTED GRAPH MATCHING PROBLEM. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 15. pp. 522-525.
language_invalid_str_mv en
network_acronym_str KFUPM
network_name_str King Fahd University of Petroleum and Minerals
oai_identifier_str oai::1985
publishDate 2020
publisher.none.fl_str_mv IEEE COMPUTER SOC
repository.mail.fl_str_mv
repository.name.fl_str_mv
repository_id_str
spelling A LINEAR-PROGRAMMING APPROACH FOR THE WEIGHTED GRAPH MATCHING PROBLEMAl-Mohamad, HADuffuaa, S. O.unknownSystemsA linear programming (LP) approach is proposed for the weighted graph matching problem. A linear program is obtained by formulating the graph matching problem in L1 norm and then transforming the resulting quadratic optimization problem to a linear one. The linear program is solved using a Simplex-based algorithm. Then, approximate 0-1 integer solutions are obtained by applying the Hungarian method on the real solutions of the linear program. The complexity of the proposed algorithm is polynomial time, and it is O(n6 L) for matching graphs of size n. The developed algorithm is compared to two other algorithms. One is based on an eigen decomposition approach and the other on a symmetric polynomial transform. Experimental results showed that the LP approach is superior in matching graphs than both other methods.IEEE COMPUTER SOCArticlePeerReviewedinfo:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/articleapplication/pdfhttps://eprints.kfupm.edu.sa/id/eprint/1985/1/a_linear_programming_approach_for_the_we_almohamad_isi_a1993lb47000012.pdf A LINEAR-PROGRAMMING APPROACH FOR THE WEIGHTED GRAPH MATCHING PROBLEM. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 15. pp. 522-525. enhttps://eprints.kfupm.edu.sa/id/eprint/1985/http://isi.kfupm.edu.sa/journals/pdf/A/a_linear_programming_approach_for_the_we_almohamad_isi_a1993lb47000012.pdf2020info:eu-repo/semantics/openAccessoai::19852019-11-01T13:30:05Z
spellingShingle A LINEAR-PROGRAMMING APPROACH FOR THE WEIGHTED GRAPH MATCHING PROBLEM
Al-Mohamad, HA
Systems
status_str publishedVersion
title A LINEAR-PROGRAMMING APPROACH FOR THE WEIGHTED GRAPH MATCHING PROBLEM
title_full A LINEAR-PROGRAMMING APPROACH FOR THE WEIGHTED GRAPH MATCHING PROBLEM
title_fullStr A LINEAR-PROGRAMMING APPROACH FOR THE WEIGHTED GRAPH MATCHING PROBLEM
title_full_unstemmed A LINEAR-PROGRAMMING APPROACH FOR THE WEIGHTED GRAPH MATCHING PROBLEM
title_short A LINEAR-PROGRAMMING APPROACH FOR THE WEIGHTED GRAPH MATCHING PROBLEM
title_sort A LINEAR-PROGRAMMING APPROACH FOR THE WEIGHTED GRAPH MATCHING PROBLEM
topic Systems
url https://eprints.kfupm.edu.sa/id/eprint/1985/1/a_linear_programming_approach_for_the_we_almohamad_isi_a1993lb47000012.pdf