Improvement And Application Of Gene Expression Programming

In 2001,Ferreira proposed a gene expression programming(GEP)algorithm by studying the genetic expression rules in biology.GEP is a new evolutionary computing method.It combines the advantages of genetic algorithm(GA)and genetic programming(GP).It uses simple coding technology to solve complex practical problems.Its advantages are mainly reflected in the form of individual expression and the processing of results.This paper describes the key technologies of GEP,including the gene and chromosome composition of GEP,the genetic operation of GEP and the fitness function of GEP.Compared with GA and GP,GEP chromosome is simple,linear,compact and easy to carry out genetic operation.In order to keep the diversity and global convergence of the population,and to solve the problem of "premature" easily,the classical GEP Algorithm is improved.The GEP genetic operator and the parameter estimation method of the function in GEP are improved respectively.The least square method is used to estimate the parameters in the function.The main work of this paper includes the following two aspects.Firstly,GEP is applied to the problem of generating test paper for the first time,and the problem of generating test paper is analyzed and simulated.The simulation results show that it is feasible to apply GEP to the problem of generating test paper.Secondly,the function mining problem is summarized and analyzed,GEP is improved,and the GEP before and after improvement is applied to function mining for simulation.The simulation results show that the improved GEP has obvious advantages in success rate and iteration number compared with the traditional GEP.