基于RBF神经网络的Web分类挖掘研究

Research of Web Classification Mining Based on RBF Neural Network

根据Web文档分类与人工神经网络理论,设计了一个Web分类挖掘系统。针对BP网络分类器的不足,提出了用径向基函数神经网络对Web页面中的文本信息进行分类的方法。实验初步证明,用径向基函数进行分类比BP算法构造的神经网络更具准确性,有效地提高了分类的正确率。

According to the theory of Web document classification and artificial neural network, we designed a Web classification mining system. BP network has lots of disadvantages, so we proposed a method that uses RBFNN to classify the text information in Web pages. Experiment verified it is more efficient to use RBF to classify than BP algorithm and promote the correct rate of classification.