Structured population genetic algorithms: a literature survey |
| |
Authors: | Ting Yee Lim |
| |
Affiliation: | 1. Universiti Sains Malaysia (USM), 11800, Penang, Malaysia
|
| |
Abstract: | The Genetic Algorithm (GA) has been one of the most studied topics in evolutionary algorithm literature. Mimicking natural processes of inheritance, mutation, natural selection and genetic operators, GAs have been successful in solving various optimization problems. However, standard GA is often criticized as being too biased in candidate solutions due to genetic drift in search. As a result, GAs sometimes converge on premature solutions. In this paper, we survey the major advances in GA, particularly in relation to the class of structured population GAs, where better exploration and exploitation of the search space is accomplished by controlling interactions among individuals in the population pool. They can be classified as spatial segregation, spatial distance and heterogeneous population. Additionally, secondary factors such as aging, social behaviour, and so forth further guide and shape the reproduction process. Restricting randomness in reproduction has been seen to have positive effects on GAs. It is our hope that by reviewing the many existing algorithms, we shall see even better algorithms being developed. |
| |
Keywords: | |
本文献已被 SpringerLink 等数据库收录! |
|