Metadata language
Biologically inspired methods of an evolutionary algorithm control
Subtitle:Raport Badawczy = Research Report ; RB/22/2001
Creator: Publisher:Instytut Badań Systemowych. Polska Akademia Nauk ; Systems Research Institute. Polish Academy of Sciences
Place of publishing: Date issued/created: Description:27 pages ; 21 cm ; Bibliography p. 26-27
Subject and Keywords:Genetic algorithms ; Algorytmy genetyczne ; Adaptation ; Adaptive evolutionary algorithms ; Adaptacyjne algorytmy ewolucyjne ; Adaptacja
Abstract:In this paper two new methods of an evolutionary algorithm control are proposed. Methods applied in the evolutionary algorithms are usually derived from nature and prefer solutions where the main role plays randomness, competition and fight among individuals. In the case of evolutionary algorithms, where populations of individuals are usually small it causes a premature convergence to local minima. To avoid this drawback we propose to apply an approach based rather on an agricultural technique. The correctness of such assumption follows from the observation that by operating on small populations of plants or animals it was possible to cultivate species of desired features without randomness, fight and competition. Two new methods of object selections are proposed: a histogram selection and a mixed selection. Also advantages of passing them into the evolutionary algorithm are shown, using examples based on scheduling and TSP.
Relation:Raport Badawczy = Research Report
Resource type: Detailed Resource Type: Source: Language: Language of abstract: Rights:Creative Commons Attribution BY 4.0 license
Terms of use:Copyright-protected material. [CC BY 4.0] May be used within the scope specified in Creative Commons Attribution BY 4.0 license, full text available at: ; -
Digitizing institution:Systems Research Institute of the Polish Academy of Sciences
Original in:Library of Systems Research Institute PAS
Projects co-financed by: Access: