A Guided Mutation Operator for Dynamic Diversity Enhancement in Evolutionary Strategies

José L. Guerrero, Antonio Berlanga and José M. Molina
Additional contact information
José L. Guerrero: Computer Science Department, Applied Artificial Intelligence Research Group, University Carlos III of Madrid, Colmenarejo, Spain
Antonio Berlanga: Computer Science Department, Applied Artificial Intelligence Research Group, University Carlos III of Madrid, Colmenarejo, Spain
José M. Molina: Computer Science Department, Applied Artificial Intelligence Research Group, University Carlos III of Madrid, Colmenarejo, Spain

International Journal of Natural Computing Research (IJNCR), 2014, vol. 4, issue 2, 20-39

Abstract: Diversity in evolutionary algorithms is a critical issue related to the performance obtained during the search process and strongly linked to convergence issues. The lack of the required diversity has been traditionally linked to problematic situations such as early stopping in the presence of local optima (usually faced when the number of individuals in the population is insufficient to deal with the search space). Current proposal introduces a guided mutation operator to cope with these diversity issues, introducing tracking mechanisms of the search space in order to feed the required information to this mutation operator. The objective of the proposed mutation operator is to guarantee a certain degree of coverage over the search space before the algorithm is stopped, attempting to prevent early convergence, which may be introduced by the lack of population diversity. A dynamic mechanism is included in order to determine, in execution time, the degree of application of the technique, adapting the number of cycles when the technique is applied. The results have been tested over a dataset of ten standard single objective functions with different characteristics regarding dimensionality, presence of multiple local optima, search space range and three different dimensionality values, 30D, 300D and 1000D. Thirty different runs have been performed in order to cover the effect of the introduced operator and the statistical relevance of the measured results

Date: 2014
References: Add references at CitEc
Citations:

Downloads: (external link)
http://services.igi-global.com/resolvedoi/resolve. ... 018/ijncr.2014040102 (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:igg:jncr00:v:4:y:2014:i:2:p:20-39

Access Statistics for this article

International Journal of Natural Computing Research (IJNCR) is currently edited by Xuewen Xia

More articles in International Journal of Natural Computing Research (IJNCR) from IGI Global
Bibliographic data for series maintained by Journal Editor ().