Linkage Learning Genetic Algorithm

Abstract

In order to handle linkage evolution and to tackle the ordering problem, Harik [47] took Holland’s call [53] for the evolution of tight linkage quite literally and proposed the linkage learning genetic algorithm (LLGA), which is capable of learning genetic linkage in the evolutionary process. The linkage learning genetic algorithm used a special probabilistic expression mechanism and a unique combination of the (gene number, allele) coding scheme and an exchange crossover operator to create an evolvable genotypic structure that made genetic linkage learning natural and viable for genetic algorithms. As the subject of this study, the design, works, accomplishments, and limitations of the linkage learning genetic algorithm are presented and discussed in this chapter. Detailed background and comprehensive description can also be found elsewhere [47, 49, 50].