基于遗传算法的BP神经网络改进

An improved BP neural network algorithm based on genetic algorithm

摘要: BP神经网络能以任意精度逼近非线性函数，但是存在收敛速度慢、易陷入局部极小值的缺点.而遗传算法具有兼顾全局和局部搜索的优点.提出了用遗传算法改进BP神经网络的一个方案，通过仿真实验的结果对比，得出了改进后的算法对数据预测的收敛速度更快、精度更高的结论.

Abstract: BP neural network can approximate nonlinear functions with any precision,but it may cause slow convergence and is more likely to get local minimums.On the contrary,genetic algorithm has strong capability of both global and local search.In this paper,a scheme is proposed,using genetic algorithm to improve BP neural network.Comparing the simulation results,a conclusion is gained that the improved algorithm converges faster and is more accurate when used to predict the result.