文本分类中一种基于正交变换的特征降维方法

A Method of Reducing Features Based on Orthogonal Transforming in Text Classification

本文讨论了一种基于正交变换的文本特征降维方法。分析了基于特征选择和特征抽取的特征降维方法各自特点,借助矩阵的分解论证了基于Fisher准则函数的特征降维模式的原理与理论基础,讨论了PCA与SVD两种模式的相互关系。实验结果表明这种特征降维模式在文本分类的准确性方面效果较好。

Discuss a method of reducing the text features based on orthogonal transforming. Then, we analyze the characteristic about feature selection and feature extraction. In virtue of decomposing the matrix, we demonstrate the principle and the theory foundation about the reduce of dimensions based on Fisher. We discuss the relation of PCA and SVD. The test shows that this method has a better precision in the text categorization.