Mid Sweden University

miun.sePublikasjoner
Endre søk
RefereraExporteraLink to record
Permanent link

Direct link
Referera
Referensformat
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Annet format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annet språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf
An experimental study on robust parameter settings
Mittuniversitetet, Fakulteten för naturvetenskap, teknik och medier, Institutionen för informationsteknologi och medier.ORCID-id: 0000-0001-9372-3416
2010 (engelsk)Inngår i: Proceedings of the 12th annual conference comp on Genetic and evolutionary computation, ACM Press, 2010, s. 1999-2002Konferansepaper, Publicerat paper (Fagfellevurdert)
Abstract [en]

That there is no best initial parameter setting for a metaheuristicon all optimization problems is a proven fact (nofree lunch theorem). This paper studies the applicability ofso called robust parameter settings for combinatorial optimizationproblems. Design of Experiments supported parameterscreening had been carried out, analyzing a discreteParticle Swarm Optimization algorithm on three demographicallyvery dissimilar instances of the Traveling SalesmenProblem. First experimental results indicate that parametersettings produce varying performance quality forthe three instances. The robust parameter setting is outperformedin two out of three cases. The results are evensignicantly worse when considering quality/time trade-o.A methodology for problem generalization is referred to asa possible solution.

sted, utgiver, år, opplag, sider
ACM Press, 2010. s. 1999-2002
Emneord [en]
Experimental Design, Metaheuristics, Parameter Tuning
HSV kategori
Identifikatorer
URN: urn:nbn:se:miun:diva-11892DOI: 10.1145/1830761.1830844ISI: 000322071400073Scopus ID: 2-s2.0-77955953052ISBN: 978-1-4503-0073-5 (tryckt)OAI: oai:DiVA.org:miun-11892DiVA, id: diva2:332053
Konferanse
GECCO - Genetic And Evolutionary Computation Conference - 2010
Tilgjengelig fra: 2010-08-02 Laget: 2010-08-01 Sist oppdatert: 2025-09-25bibliografisk kontrollert
Inngår i avhandling
1. Automatic Instance-based Tailoring of Parameter Settings for Metaheuristics
Åpne denne publikasjonen i ny fane eller vindu >>Automatic Instance-based Tailoring of Parameter Settings for Metaheuristics
2011 (engelsk)Licentiatavhandling, med artikler (Annet vitenskapelig)
Abstract [en]

Many industrial problems in various fields, such as logistics, process management, orproduct design, can be formalized and expressed as optimization problems in order tomake them solvable by optimization algorithms. However, solvers that guarantee thefinding of optimal solutions (complete) can in practice be unacceptably slow. Thisis one of the reasons why approximative (incomplete) algorithms, producing near-optimal solutions under restrictions (most dominant time), are of vital importance.

Those approximative algorithms go under the umbrella term metaheuristics, each of which is more or less suitable for particular optimization problems. These algorithmsare flexible solvers that only require a representation for solutions and an evaluation function when searching the solution space for optimality.What all metaheuristics have in common is that their search is guided by certain control parameters. These parameters have to be manually set by the user andare generally problem and interdependent: A setting producing near-optimal resultsfor one problem is likely to perform worse for another. Automating the parameter setting process in a sophisticated, computationally cheap, and statistically reliable way is challenging and a significant amount of attention in the artificial intelligence and operational research communities. This activity has not yet produced any major breakthroughs concerning the utilization of problem instance knowledge or the employment of dynamic algorithm configuration.

The thesis promotes automated parameter optimization with reference to the inverse impact of problem instance diversity on the quality of parameter settings with respect to instance-algorithm pairs. It further emphasizes the similarities between static and dynamic algorithm configuration and related problems in order to show how they relate to each other. It further proposes two frameworks for instance-based algorithm configuration and evaluates the experimental results. The first is a recommender system for static configurations, combining experimental design and machine learning. The second framework can be used for static or dynamic configuration,taking advantage of the iterative nature of population-based algorithms, which is a very important sub-class of metaheuristics.

A straightforward implementation of framework one did not result in the expected improvements, supposedly because of pre-stabilization issues. The second approach shows competitive results in the scenario when compared to a state-of-the-art model-free configurator, reducing the training time by in excess of two orders of magnitude.

sted, utgiver, år, opplag, sider
Östersund: Mid Sweden University, 2011. s. 62
Serie
Mid Sweden University licentiate thesis, ISSN 1652-8948 ; 67
Emneord
Algorithm Configuration, Parameter Tuning, Parameter Control, Metaheuristics
HSV kategori
Identifikatorer
urn:nbn:se:miun:diva-14613 (URN)978-91-86694-48-7 (ISBN)
Presentation
2011-10-14, Q221, Akademigatan 1, Östersund, 22:41 (engelsk)
Opponent
Veileder
Tilgjengelig fra: 2011-10-17 Laget: 2011-10-16 Sist oppdatert: 2025-09-25bibliografisk kontrollert

Open Access i DiVA

Fulltekst mangler i DiVA

Andre lenker

Forlagets fulltekstScopus

Person

Dobslaw, Felix

Søk i DiVA

Av forfatter/redaktør
Dobslaw, Felix
Av organisasjonen

Søk utenfor DiVA

GoogleGoogle Scholar

doi
isbn
urn-nbn

Altmetric

doi
isbn
urn-nbn
Totalt: 655 treff
RefereraExporteraLink to record
Permanent link

Direct link
Referera
Referensformat
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Annet format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annet språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf