一种可修正激励函数的BP算法

A BP Algorithm with Corrected Activation Function

以往的ＢＰ算法调节神经元网络的权值，其网络的隐层结点数、网络学习快慢程度及网络的泛化能力都与网络的激励函数有关的。为此，本文提出了一种带可以修正激励函数的ＢＰ算法，其特点是它能更好地模拟人脑神经元的特性。通过仿真验证此方法是非常有效的。

BP algorithm is often used to correct weights of neural network because number of hidden nodes,studying speed and generation ability of neural network are related to activation function.This paper presents a BP algorithm with corrected activation function.Its advantage is that it can better simulate neuron of human being.Practical example indicates that the method works well.