Double System Co-Evolutionary Gene Expression Programming and Its Application in Function Finding Problems

Inspired by the evolution process of human, a double system co-evolutionary gene expression programming (DSCEGEP) is proposed in this paper. The DSCE-GEP consists of natural evolution system and artificial intervention system, where the natural evolution system solves problem according to the mode of standard gene expression programming (STGEP) and the artificial intervention system guides the process of natural evolution system to solve problem. The artificial intervention system includes individual intervention operation and population intervention operation and is based on a gene pool, which is first created according to prior knowledge and then is updated according to posterior knowledge, contains eminent genes, morbid genes and their characteristic information as well. The individual intervention operation, which aims to improve the quality of population individuals, consists of a repairing operator that removes the morbid genes in individuals and a strengthening operator that spreads eminent genes to the individuals of population. The population intervention operation uses entropy as the diversity indicators of population and increases the population entropy by introducing a certain number of feasible random individuals and feasible mirror individuals to improve the global searching ability of the algorithm when the population entropy drops to a given threshold. The experimental results show that the performance of DSCE-GEP is better than other GEP algorithms proposed in the related literatures as regards function finding problems, and promises competitive performance.