25th International Symposium on Multiple-Valued Logic Some New Results for Multiple-Valued Genetic Algorithms Bloomington, Indiana May 23-May 25 ISBN: 0-8186-7118-1
Abstract: The paper describes each of the operations involved in a genetic algorithm: reproduction, mutation, and selection, and discusses each in the language of classical multiple-valued logic. The differences among forms of reproduction that have been used by various researchers are examined and the relative importance of each of the operations in searching for highly fit members of a population is evaluated. The role of mutation in ensuring the completeness of the set of genetic operators is established. A recently proposed form of selection is shown to force convergence of the genetic algorithm, independently of reproduction and mutation. Finally, the theorems developed are applied to practical problems in the use of genetic algorithms.
Index Terms:
genetic algorithms; multivalued logic; convergence of numerical methods; multiple-valued genetic algorithms; selection; mutation; reproduction; classical multiple-valued logic; highly fit population members; genetic operators; convergence; theorems
Citation:
T.C. Wesselkamper, J. Danowitz, "Some New Results for Multiple-Valued Genetic Algorithms," ismvl, pp.0264, 25th International Symposium on Multiple-Valued Logic, 1995 Usage of this product signifies your acceptance of the Terms of Use. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Digg
Furl
Spurl
Blink
Simpy
Google
Del.icio.us
Y!MyWeb