Redundant representations in evolutionary computation
- Publication type:
- Journal article
- Metadata:
-
- Autoren
- Franz Rothlauf
- David E. Goldberg
- Sammlungen
- metadata
- ISSN
- 1063-6560
- Ausgabe der Veröffentlichung
- 4
- Zeitschrift
- Evolutionary computation
- Schlüsselwörter
- 004 Informatik
- 004 Data processing
- Sprache
- eng
- Paginierung
- Seiten: 381 - 415
- Datum der Veröffentlichung
- 2003
- Herausgeber
- MIT Press
- Datum der Datenerfassung
- 2020
- Datum, an dem der Datensatz öffentlich gemacht wurde
- 2020
- Zugang
- Public
- Titel
- Redundant representations in evolutionary computation
- Ausgabe der Zeitschrift
- 11
Data source: METADATA.UB
- Other metadata sources:
-
- Autoren
- F Rothlauf
- DE Goldberg
- Autoren-URL
- https://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=fis-test-1&SrcAuth=WosAPI&KeyUT=WOS:000186514400004&DestLinkType=FullRecord&DestApp=WOS_CPL
- DOI
- 10.1162/106365603322519288
- eISSN
- 1530-9304
- Externe Identifier
- Clarivate Analytics Document Solution ID: 742JV
- PubMed Identifier: 14629864
- ISSN
- 1063-6560
- Ausgabe der Veröffentlichung
- 4
- Zeitschrift
- EVOLUTIONARY COMPUTATION
- Schlüsselwörter
- redundant representations
- neutral theory
- neutral networks
- synonymously redundant
- non-synonymously redundant
- link-and-node biased encoding
- trivial voting mapping
- Paginierung
- 381 - 415
- Datum der Veröffentlichung
- 2003
- Status
- Published
- Titel
- Redundant representations in evolutionary computation
- Sub types
- Article
- Ausgabe der Zeitschrift
- 11
Data source: Web of Science (Lite)
- Abstract
- <jats:p> This paper discusses how the use of redundant representations influences the performance of genetic and evolutionary algorithms. Representations are redundant if the number of genotypes exceeds the number of phenotypes. A distinction is made between synonymously and non-synonymously redundant representations. Representations are synonymously redundant if the genotypes that represent the same phenotype are very similar to each other. Non-synonymously redundant representations do not allow genetic operators to work properly and result in a lower performance of evolutionary search. When using synonymously redundant representations, the performance of selectorecombinative genetic algorithms (GAs) depends on the modification of the initial supply. We have developed theoretical models for synonymously redundant representations that show the necessary population size to solve a problem and the number of generations goes with O(2<jats:sup>k</jats:sup><jats:sup>r</jats:sup>/r), where k<jats:sub>r</jats:sub> is the order of redundancy and r is the number of genotypic building blocks (BB) that represent the optimal phenotypic BB. As a result, uniformly redundant representations do not change the behavior of GAs. Only by increasing r, which means overrepresenting the optimal solution, does GA performance increase. Therefore, non-uniformly redundant representations can only be used advantageously if a-priori information exists regarding the optimal solution. The validity of the proposed theoretical concepts is illustrated for the binary trivial voting mapping and the real-valued link-biased encoding. Our empirical investigations show that the developed population sizing and time to convergence models allow an accurate prediction of the empirical results. </jats:p>
- Autoren
- Franz Rothlauf
- David E Goldberg
- DOI
- 10.1162/106365603322519288
- eISSN
- 1530-9304
- ISSN
- 1063-6560
- Ausgabe der Veröffentlichung
- 4
- Zeitschrift
- Evolutionary Computation
- Sprache
- en
- Paginierung
- 381 - 415
- Datum der Veröffentlichung
- 2003
- Status
- Published
- Herausgeber
- MIT Press - Journals
- Herausgeber URL
- http://dx.doi.org/10.1162/106365603322519288
- Datum der Datenerfassung
- 2021
- Titel
- Redundant Representations in Evolutionary Computation
- Ausgabe der Zeitschrift
- 11
Data source: Crossref
- Abstract
- This paper discusses how the use of redundant representations influences the performance of genetic and evolutionary algorithms. Representations are redundant if the number of genotypes exceeds the number of phenotypes. A distinction is made between synonymously and non-synonymously redundant representations. Representations are synonymously redundant if the genotypes that represent the same phenotype are very similar to each other. Non-synonymously redundant representations do not allow genetic operators to work properly and result in a lower performance of evolutionary search. When using synonymously redundant representations, the performance of selectorecombinative genetic algorithms (GAs) depends on the modification of the initial supply. We have developed theoretical models for synonymously redundant representations that show the necessary population size to solve a problem and the number of generations goes with O(2(kr)/r), where kr is the order of redundancy and r is the number of genotypic building blocks (BB) that represent the optimal phenotypic BB. As a result, uniformly redundant representations do not change the behavior of GAs. Only by increasing r, which means overrepresenting the optimal solution, does GA performance increase. Therefore, non-uniformly redundant representations can only be used advantageously if a-priori information exists regarding the optimal solution. The validity of the proposed theoretical concepts is illustrated for the binary trivial voting mapping and the real-valued link-biased encoding. Our empirical investigations show that the developed population sizing and time to convergence models allow an accurate prediction of the empirical results.
- Addresses
- Department of Information Systems 1, University of Mannheim, Schloss, D-68131 Mannheim, Germany. rothlauf@uni-mannheim.de
- Autoren
- Franz Rothlauf
- David E Goldberg
- DOI
- 10.1162/106365603322519288
- eISSN
- 1530-9304
- Externe Identifier
- PubMed Identifier: 14629864
- Open access
- false
- ISSN
- 1063-6560
- Ausgabe der Veröffentlichung
- 4
- Zeitschrift
- Evolutionary computation
- Schlüsselwörter
- Computational Biology
- Population Density
- Genotype
- Phenotype
- Algorithms
- Models, Genetic
- Time Factors
- Biological Evolution
- Sprache
- eng
- Medium
- Paginierung
- 381 - 415
- Datum der Veröffentlichung
- 2003
- Status
- Published
- Datum der Datenerfassung
- 2003
- Titel
- Redundant representations in evolutionary computation.
- Sub types
- Journal Article
- Ausgabe der Zeitschrift
- 11
Data source: Europe PubMed Central
- Abstract
- This paper discusses how the use of redundant representations influences the performance of genetic and evolutionary algorithms. Representations are redundant if the number of genotypes exceeds the number of phenotypes. A distinction is made between synonymously and non-synonymously redundant representations. Representations are synonymously redundant if the genotypes that represent the same phenotype are very similar to each other. Non-synonymously redundant representations do not allow genetic operators to work properly and result in a lower performance of evolutionary search. When using synonymously redundant representations, the performance of selectorecombinative genetic algorithms (GAs) depends on the modification of the initial supply. We have developed theoretical models for synonymously redundant representations that show the necessary population size to solve a problem and the number of generations goes with O(2(kr)/r), where kr is the order of redundancy and r is the number of genotypic building blocks (BB) that represent the optimal phenotypic BB. As a result, uniformly redundant representations do not change the behavior of GAs. Only by increasing r, which means overrepresenting the optimal solution, does GA performance increase. Therefore, non-uniformly redundant representations can only be used advantageously if a-priori information exists regarding the optimal solution. The validity of the proposed theoretical concepts is illustrated for the binary trivial voting mapping and the real-valued link-biased encoding. Our empirical investigations show that the developed population sizing and time to convergence models allow an accurate prediction of the empirical results.
- Autoren
- Franz Rothlauf
- David E Goldberg
- Autoren-URL
- https://www.ncbi.nlm.nih.gov/pubmed/14629864
- DOI
- 10.1162/106365603322519288
- ISSN
- 1063-6560
- Ausgabe der Veröffentlichung
- 4
- Zeitschrift
- Evol Comput
- Schlüsselwörter
- Algorithms
- Biological Evolution
- Computational Biology
- Genotype
- Models, Genetic
- Phenotype
- Population Density
- Time Factors
- Sprache
- eng
- Country
- United States
- Paginierung
- 381 - 415
- Datum der Veröffentlichung
- 2003
- Status
- Published
- Datum, an dem der Datensatz öffentlich gemacht wurde
- 2004
- Titel
- Redundant representations in evolutionary computation.
- Sub types
- Journal Article
- Ausgabe der Zeitschrift
- 11
Data source: PubMed
- Autoren
- Franz Rothlauf
- David E Goldberg
- DOI
- 10.1162/106365603322519288
- Zeitschrift
- Evol. Comput.
- Artikelnummer
- 4
- Paginierung
- 381 - 415
- Datum der Veröffentlichung
- 2003
- Titel
- Redundant Representations in Evolutionary Computation.
- Ausgabe der Zeitschrift
- 11
Data source: DBLP
- Autoren
- Ausgabe der Veröffentlichung
- 4
- Zeitschrift
- Evolutionary Computation
- Paginierung
- 381 - 415
- Datum der Veröffentlichung
- 2003
- Titel
- Redundant representations in evolutionary computation
- Ausgabe der Zeitschrift
- 11
Data source: CiNii EN
- Beziehungen:
- Property of