一种基于反馈学习机制的支持向量机网页分类方法

Research on Webpage Classification Based on Feedback SVM

本文以SVM分类方法为基础,研究了SVM在网页分类方面的应用并给出了基于KKT条件的反馈机制对SVM方法进行改进的方法。通过对中等规模的Web网页测试实验表明基于KKT(karush—Kuhn—Tucker)条件的反馈学习机制的支持向量机对Web网页分类方法是有效的。

Based on the SVM classification method, this paper studied on feedback SVM method for the Webpage classification. This paper gives an improved method based on KKT conditions. Experimental results on a middle scale webpage show that the method is effective.