A Hierarchical Model of Parallel Genetic Programming Applied to Bioinformatic Problems

摘要

Genetic Programming (GP), an evolutionary method, can be used to solve difficult problems in various applications. However, three important problems in GP are its tendency to find non-parsimonious solutions (bloat), to converge prematurely and to use a tremendous amount of computing time. In this paper, we present an efficient model of distributed GP to limit these general GP drawbacks. This model uses a multi-objective optimization and a hierarchical communication topology.