Title :
Crossover interactions among niches
Author :
Mahfoud, Samir W.
Author_Institution :
Dept. of Comput. Sci., Illinois Univ., Urbana, IL, USA
Abstract :
Niching methods in genetic algorithms (GAs) strive to locate and maintain stable subpopulations or niches, within a single population. This paper explores the distribution of population elements among niches arising under deterministic crowding. Crossover interactions among niches are observed, isolated, and explained. Ideal behavior is compared with actual behavior on a variety of multimodal optimization problems, including a multimodal deceptive problem
Keywords :
genetic algorithms; optimisation; crossover interactions; deterministic crowding; genetic algorithms; multimodal deceptive problem; multimodal optimization problems; niches; niching methods; single population; subpopulations; Computer science; Contracts; DC generators; Genetic algorithms; Genetic mutations; Hamming distance; Stochastic processes;
Conference_Titel :
Evolutionary Computation, 1994. IEEE World Congress on Computational Intelligence., Proceedings of the First IEEE Conference on
Conference_Location :
Orlando, FL
Print_ISBN :
0-7803-1899-4
DOI :
10.1109/ICEC.1994.350018
