基于小波网络和径向基网络的分类系统

Classification System Based on Wavelet Network & RBF Network

文章在分析小波函数与径向基函数关系的基础上,构造了一个神经网络联合分类系统。该系统主要由分类层、决策层以及信息反馈与交换模块构成。分类层由小波网络和RBF网络构成,整个系统的联合决策规则采用SUM规则,信息反馈与交换模块的工作原理是基于小波函数与径向基函数的关系。我们对系统各部分做了分析,并对系统整体性能做了测试。

In this paper by analyzing the relationship between wavelet and RBF, we construct a classification system based on neural networks. This system consists of classification part, decision part and a module for feed-backing and changing information. The classification part consists of wavelet network and RBF network, the whole system's decision combination rule is SUM rule, and the working theory of the module is based on the relationship between wavelet and RBF. We analyze the three parts of the system and test the performance of the whole system.