Title :
Shuffle crossover and mutual information
Author :
Burkowski, Forbes J.
Author_Institution :
Dept. of Comput. Sci., Waterloo Univ., Ont., Canada
Abstract :
We introduce a crossover operator that is not dependent on the initial layout of the genome. While maintaining a low positional bias, the MISC (mutual information and shuffle crossover) algorithm is competitive with one-point crossover and works by automatically regrouping bits that are considered to be interdependent. The heuristic strategy is to derive an operator that promotes the expansion of building blocks as required by a genetic algorithm
Keywords :
genetic algorithms; heuristic programming; information theory; mathematical operators; MISC algorithm; automatic bit regrouping; crossover operator; genetic algorithm; genome; heuristic strategy; low positional bias; mutual information; mutual information and shuffle crossover algorithm; one-point crossover; shuffle crossover; Bayesian methods; Bioinformatics; Character generation; Computer science; Genetic algorithms; Genetic communication; Genomics; Ground penetrating radar; Information theory; Mutual information;
Conference_Titel :
Evolutionary Computation, 1999. CEC 99. Proceedings of the 1999 Congress on
Conference_Location :
Washington, DC
Print_ISBN :
0-7803-5536-9
DOI :
10.1109/CEC.1999.782671
