Scalability of using Restricted Boltzmann Machines for combinatorial optimization

Publication type:
Journal article
Metadata:
Autoren
Sammlungen
  • metadata
ISSN
0377-2217
Ausgabe der Veröffentlichung
2
Zeitschrift
European journal of operational research
Schlüsselwörter
  • 330 Wirtschaft
  • 330 Economics
Sprache
eng
Paginierung
Seiten: 368 - 383
Datum der Veröffentlichung
2017
Herausgeber
Elsevier
Herausgeber URL
http://dx.doi.org/10.1016/j.ejor.2016.06.066
Datum der Datenerfassung
2020
Datum, an dem der Datensatz öffentlich gemacht wurde
2020
Zugang
Public
Titel
Scalability of using Restricted Boltzmann Machines for combinatorial optimization
Ausgabe der Zeitschrift
256

Data source: METADATA.UB

Other metadata sources:
Autoren
Autoren-URL
https://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=fis-test-1&SrcAuth=WosAPI&KeyUT=WOS:000384857500003&DestLinkType=FullRecord&DestApp=WOS_CPL
DOI
10.1016/j.ejor.2016.06.066
eISSN
1872-6860
Externe Identifier
  • Clarivate Analytics Document Solution ID: DY1LW
ISSN
0377-2217
Ausgabe der Veröffentlichung
2
Zeitschrift
EUROPEAN JOURNAL OF OPERATIONAL RESEARCH
Schlüsselwörter
  • Combinatorial optimization
  • Heuristics
  • Evolutionary computation
  • Estimation of Distribution Algorithms
  • Neural Networks
Paginierung
368 - 383
Datum der Veröffentlichung
2017
Status
Published
Titel
Scalability of using Restricted Boltzmann Machines for combinatorial optimization
Sub types
  • Article
Ausgabe der Zeitschrift
256

Data source: Web of Science (Lite)

Autoren
DOI
10.1016/j.ejor.2016.06.066
ISSN
0377-2217
Ausgabe der Veröffentlichung
2
Zeitschrift
European Journal of Operational Research
Sprache
en
Paginierung
368 - 383
Datum der Veröffentlichung
2017
Status
Published
Herausgeber
Elsevier BV
Herausgeber URL
http://dx.doi.org/10.1016/j.ejor.2016.06.066
Datum der Datenerfassung
2019
Titel
Scalability of using Restricted Boltzmann Machines for combinatorial optimization
Ausgabe der Zeitschrift
256

Data source: Crossref

Autoren
Zeitschrift
CoRR
Datum der Veröffentlichung
2014
Titel
Scalability of using Restricted Boltzmann Machines for Combinatorial Optimization.
Ausgabe der Zeitschrift
abs/1411.7542

Data source: DBLP

Abstract
Estimation of Distribution Algorithms (EDAs) require flexible probability models that can be efficiently learned and sampled. Restricted Boltzmann Machines (RBMs) are generative neural networks with these desired properties. We integrate an RBM into an EDA and evaluate the performance of this system in solving combinatorial optimization problems with a single objective. We assess how the number of fitness evaluations and the CPU time scale with problem size and with problem complexity. The results are compared to the Bayesian Optimization Algorithm, a state-of-the-art EDA. Although RBM-EDA requires larger population sizes and a larger number of fitness evaluations, it outperforms BOA in terms of CPU times, in particular if the problem is large or complex. RBM-EDA requires less time for model building than BOA. These results highlight the potential of using generative neural networks for combinatorial optimization.
Autoren
Autoren-URL
http://arxiv.org/abs/1411.7542v1
Schlüsselwörter
  • cs.NE
  • cs.NE
  • I.2.6; I.2.8
Datum der Veröffentlichung
2014
Datum der Datenerfassung
2014
Datum, an dem der Datensatz öffentlich gemacht wurde
2014
Titel
Scalability of using Restricted Boltzmann Machines for Combinatorial Optimization

Files

1411.7542v1.pdf

Data source: arXiv

Abstract
Estimation of Distribution Algorithms (EDAs) require flexible probability models that can be efficiently learned and sampled. Restricted Boltzmann Machines (RBMs) are generative neural networks with these desired properties. We integrate an RBM into an EDA and evaluate the performance of this system in solving combinatorial optimization problems with a single objective. We assess how the number of fitness evaluations and the CPU time scale with problem size and complexity. The results are compared to the Bayesian Optimization Algorithm (BOA), a state-of-the-art multivariate EDA, and the Dependency Tree Algorithm (DTA), which uses a simpler probability model requiring less computational effort for training the model. Although RBM–EDA requires larger population sizes and a larger number of fitness evaluations than BOA, it outperforms BOA in terms of CPU times, in particular if the problem is large or complex. This is because RBM–EDA requires less time for model building than BOA. DTA with its restricted model is a good choice for small problems but fails for larger and more difficult problems. These results highlight the potential of using generative neural networks for combinatorial optimization.
Autoren
Ausgabe der Veröffentlichung
2
Schlüsselwörter
  • Combinatorial optimization
  • Heuristics
  • Evolutionary computation
  • Estimation of Distribution Algorithms
  • Neural Networks
Paginierung
368 - 383
Datum der Veröffentlichung
2017
Titel
Scalability of using Restricted Boltzmann Machines for combinatorial optimization
Ausgabe der Zeitschrift
256

Data source: RePEc

Beziehungen:
Property of

Tools