摘要
将SVM和KNN算法结合在一起,组成一种新的Web文本分类算法——SVM-KNN算法。当Web文本和SVM最优超平面的距离大于预选设定的阈值,则采用SVM进行分类,反之采用SVM作为代表点的KNN算法对样本分类。实证结果表明,SVM-KNN分类算法的分类精度比单纯SVM或KNN分类算法有不同程度的提高,为Web数据挖掘提供了一种有效的分类方法。
This paper used SVM and KNN algorithm together to form a new classification algorithm for Web text—SVM-KNN algorithm.When optimal super plane distance of Web text and SVM was greater than the preselected threshold,used SVM to classify,otherwise it adopted KNN algorithm to classify the samples of SVM as the representative point.The experimental results show that the accuracy of SVM-KNN classification algorithm are better than pure SVM or KNN classification algorithm,and the Web text classification provides an effective classification method.
出处
《计算机应用研究》
CSCD
北大核心
2012年第5期1926-1928,共3页
Application Research of Computers
基金
教育部2009年"春晖计划"资助项目(Z2009-1-63004)
关键词
支持向量机
数据挖掘
网页分类
特征提取
support vector machines(SVM)
data mining(DM)
Web classification
feature selection