摘要
近几年来,模糊神经网络(FNN)的研究引起了广泛的注意。本文对FNN上的反向传播学习方法加以讨论。使用输入均值和输出权重参量来进行模糊化和反模糊化处理,学习的目的是调整这两个参量到合适的值。
This paper designs an FNN with five layers and discusses the problem of its back - propagation learning. It employes input - mean and output - weight to implement fuzzication and defuzzication. This algorithm aims at adjusting input - means and output - weights to more proper values.
出处
《计算机应用与软件》
CSCD
1998年第4期34-38,共5页
Computer Applications and Software
关键词
模糊神经网络
反向传播学习
算法
FNN (Fuzzy Neural Network ), back - propagation learning, input - mean, output - weight.
