基于改进的模糊聚类RBF网络集成的文本分类方法

Text classification based on improved Fuzzy clustering RBF neural network ensemble

Web文本分类是数据挖掘研究的一个热点问题.针对文本向量维度过高的特点,提出一种改进的模糊聚类RBF网络集成的文本分类方法,该方法利用模糊C均值聚类算法对文本特征向量进行简化、抽取,引入自适应遗传算法优化RBF神经网络的权值,构建RBF网络集成模型对文本进行分类.实验结果表明,该方法具有更高的分类效率和正确率.

Web text classification is a hot issue in the research on data mining. In view of the characteris- tics of high dimension text vector, the paper proposes an improved text classification method of fuzzy cluster RBF network integration. The method uses fuzzy c-means clustering algorithm to simplify and extract the text eigenvector, introduces adaptive genetic algorithm for optimization of RBF Neural net- work weights, and builds a RBF network model for text classification. Experimental results show that the method possesses a higher ela^if;o.~;.,~ ~gg;~： ,