模糊神经网络识别方法

Fuzzy Neural Network Recognition Method

本文提出了一种新的模糊神经网络分类器，它结合了神经网络和模糊系统的优点，在使用神经网络结构的同时又用模糊理论来处理模糊和非模糊的信息。新网络在处理分类问题时，对畸变和噪声有较好的容忍性。本文还给出了用网络识别数字和印鉴的例子。

In this paper, a new fuzzy neural network is proposed. The network is a combination of neural network and fuzzy system while possessing the advantages of these two systems. It has a neural network structure while using fuzzy concepts to process fuzzy and non-fuzzy information. The new fuzzy neural network has a good ability to tolerate pattern distortion and noise while processing classification problem for example number and stamp recognition.