一种改进的BP人工神经网络模型

针对人工神经网络 BP 模型在实际应用中会遇到一些问题。该文指出了 BP 神经网络模型结构和算法的一些不足,并针对 BP网络模型算法存在的局部极小问题、BP 网络学习速度较慢以及 BP 模型网络结构训练过度和降低网络归纳能力等问题提出了引入规则、选择激发函数和转换函数、采用隐含神经元以及改变网络拓扑结构等一些改进措施,对于提高 BP 模型在实际应用中的效果具有重要意义。通过实际应用,证明了改进后的算法是行之有效的。

An improved BP ANN model

There are some problems in the practice application of the artificial neural network BP model,points out some shortcomings of BP neural network model structure and of its algorithm,and puts forward the method of using rule,choosing stimulating function and transferring function,using hiding neural cell and changing the topological structure of the network to solve the problems existing in the BP neural network model structure,such as the appearance of local minimum in the neural network model algorithm,the relative slowness of the learning speed of the BP network,over-training of BP network model structure and drop of the induction performance of the BP neural network.Thus,this method is very significant in the attempt of raising the effect of the BP model in practical application of the model,and it is proved that the improved BP neural network model algorithm is very effectual in practical application.