Combining Metaheuristics with Column Generation: Successful Approaches to Enhance Column Generation Algorithms Performance

Fabian Castaño 1 Marc Sevaux 2
1 Universidad Pontificia Bolivariana
2 Lab-STICC_UBS_CACS_MOCS
Lab-STICC - Laboratoire des sciences et techniques de l'information, de la communication et de la connaissance
Abstract : Column generation algorithms are typically adopted to address mathematical programming problems defined over a huge number of variables. This approach suffers, however, from several problems that might limit its usability. In this work some of these problems are discussed along with several strategies that take advantage of the use of (meta-)heuristics to help improve the methods performance and reduce the computational effort required to compute an optimal solution. In this work the benefits of using metaheuristic strategies within CG are discussed from the viewpoint of the way they, indirectly, address some of the causes leading to a poor performance. These different methods are tested by solving the maximum network lifetime problem in wireless sensor networks for which a model that naturally leads to column generation is considered. Experimental results show how the use of metaheuristics can generate large improvements on the performance of the basic column generation framework.
Keywords : Column generation Metaheuristics Matheuristics
Document type :
Conference papers
Liste complète des métadonnées
https://hal.archives-ouvertes.fr/hal-01367946
Contributor : Marc Sevaux <>
Submitted on : Saturday, September 17, 2016 - 4:38:18 PM
Last modification on : Monday, February 25, 2019 - 3:14:12 PM

Identifiers

  • HAL Id : hal-01367946, version 1

Collections

UNIV-BREST | INSTITUT-TELECOM | LAB-STICC | ENIB | LAB-STICC_ENIB | LAB-STICC_UBS

Citation

Fabian Castaño, Marc Sevaux. Combining Metaheuristics with Column Generation: Successful Approaches to Enhance Column Generation Algorithms Performance. Proceedings of the Sixth International Workshop on Model-based Metaheuristic (Matheuristics 2016), IRIDIA/ULB, Sep 2016, Brussels, Belgium. pp.95-100. ⟨hal-01367946⟩

Export

BibTeX TEI DC DCterms EndNote

Share

Metrics

Record views

368

Contact

support.ccsd.cnrs.fr
hal.support@ccsd.cnrs.fr
Logo CCSD