Evolutionary Programming Using Mutations Based on the t Probability Distribution
M. Zhang · Dianzi xuebao · 2008
Consulted some previous mutations,this paper proposed an evolutionary programming(EP)using mutations based on the t probability distribution(tEP).The t probability distribution can connect with Gaussian and Cauchy probability distribution.Its variance is changeable by adjusting degree of freedom n.Rules of t mutations according to the change ofn and standard deviation are analyzed by simulations,and the relationship among number of generations,standard deviation and minimum fitness are briefly summarized.Then an EP using adaptive t mutations(AtEP)is put forward at the end.The AtEP uses several t mutations with different n at the same time,so it can integrate the advantages of Gaussian and Cauchy mutation.And simulation evidence shows the algorithm is better than classical EP algorithm.