Accelerating Column Generation Algorithm Using Machine-Learning-Based Column Elimination

Lichang Fang, Haofeng Yuan, Yuli Zhang, Shiji Song

    Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

    Abstract

    The column generation (CG) algorithm is widely used in large-scale optimization problems. However, a large amount of columns in the restricted master problem (RMP) makes the computing process very time-consuming. This paper proposes a machine learning based column elimination strategy to accelerate the CG algorithm. Our approach represents the RMP by a bipartite graph and applies a learned Graph Neural Network model to predict redundant columns to be eliminated from the RMP, so as to reduce the time cost of solving the RMP and iterations required for convergence. Our approach is tested on cutting stock problem instances. Compared with the vanilla CG algorithm, the iterations and time required for convergence are reduced by up to 31 % and 48%, respectively. Furthermore, our approach shows great generalization to cutting stock problem instances of different sizes.

    Original languageEnglish
    Title of host publication2023 IEEE International Conference on Systems, Man, and Cybernetics
    Subtitle of host publicationImproving the Quality of Life, SMC 2023 - Proceedings
    PublisherInstitute of Electrical and Electronics Engineers Inc.
    Pages1945-1950
    Number of pages6
    ISBN (Electronic)9798350337020
    DOIs
    Publication statusPublished - 2023
    Event2023 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2023 - Hybrid, Honolulu, United States
    Duration: 1 Oct 20234 Oct 2023

    Publication series

    NameConference Proceedings - IEEE International Conference on Systems, Man and Cybernetics
    ISSN (Print)1062-922X

    Conference

    Conference2023 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2023
    Country/TerritoryUnited States
    CityHybrid, Honolulu
    Period1/10/234/10/23

    Fingerprint

    Dive into the research topics of 'Accelerating Column Generation Algorithm Using Machine-Learning-Based Column Elimination'. Together they form a unique fingerprint.

    Cite this