Introducing Start Expression Genes to the Linkage Learning Genetic Algorithm

Abstract:

This paper discusses the use of start expression genes and a modified
exchange crossover operator in the linkage learning genetic algorithm
(LLGA) that enables the genetic algorithm to learn the linkage of
building blocks (BBs) through probabilistic expression (PE). The
difficulty that the original LLGA encounters is shown with empirical
results. Based on the observation, start expression genes and a modified
exchange crossover operator are proposed to enhance the ability of the
original LLGA to separate BBs and to improve LLGA’s performance on
uniformly scaled problems. The effect of the modifications is also
presented in the paper.

This entry was posted in Technical reports. Bookmark the permalink.
Source