Using Hierarchical Genetic Populations to Improve Solution Quality
暂无分享,去创建一个
A multi-population genetic algorithm is proposed which is hierarchical in nature. This allows the algorithm to solve problems which consist of smaller tasks contributing to the solution of an overall problem. The algorithm feeds the entire pool of individuals between local populations (solving the smaller problems) and a global population (solving the overall task). Results on categorisation tasks are presented.
[1] David Mark Levine,et al. A parallel genetic algorithm for the set partitioning problem , 1995 .