基于模糊感知器的神经模糊分类器

A Neuro-Fuzzy Classifier Based on the Fuzzy Perceptron

针对一类基于模糊感知器的神经模糊分类器 ,分析了隶属函数限制条件对分类结果的影响 .并根据模糊密度的概念 ,提出一种为输入特征选择合适的模糊划分的方法 .

For a neuro_fuzzy classifier based on the fuzzy perceptron, this paper analyses how membership function constraints affect the classification result. Based on these, according to the concept of the fuzzy density, a method is proposed to choose suitable fuzzy partitions for input features. Simulations show that it can effectively improve the classification correct rate.