A Modified Differential Evolution Algorithm for Optimization Neural Network

摘要

A Modified Differential Evolution (MDE) is proposed, which is based on the basic Differential Evolution (DE) algorithm principle and implementing framework of DE. Optimizing the initial individuals with the 1/2 rule,then by introducing the reorganization of Evolution Strategies during the period of mutation procedures.The MDE is used to optimize the weights of the feed-forward multilayer neural network, and compared with the basic DE and BP algorithm with momentum term. Finally, the numerical simulation results show that this method has good quality of high-speed global convergence and effectively improves the precision and convergence speed for feed-forward multilayer neural network. It has been proved the effectiveness and feasibility.