An Improved Gene Expression Programming for Function Finding
Gene Expression Programming (GEP) is a relative new evolutionary algorithm based on genome and phenome, and it is very effective. However, the conventional GEP may need more computation effort to solve the problems with large data because of the transformation from chromosome to expression trees. In this paper, a new method to read the gene based on Gene Expression is proposed. Furthermore, we propose a novel method to evaluate the fitness of the individual, the fitness could be computed directly without transforming the chromosome into expression trees and computing the effective length of gene so that it can reduce the computation time. To validate the efficiency of the improved GEP (IGEP), we use it to solve the function finding problems. The experiment results show that our proposed approach is not only simple and effective, but also improves the speed of computing. And it is comparable with other state-of-the-art approaches.
Gene expression programming expression trees effective length of gene
Xiaobo Liu Zhihua Cai Yuzheng Zhang
School of Computer science, China University of Geosciences, Wuhan 430074, China
国际会议
武汉
英文
2007-09-21（万方平台首次上网日期，不代表论文的发表时间）