Acta Scientiarum Naturalium Universitatis Pekinensis

Previous Articles     Next Articles

A Novel Classification Method—AMC-ASVM

SUN Xichen1,HE Renya2,FENG Jufu1   

  • Received:2006-02-02 Online:2007-01-20 Published:2007-01-20

一种新的分类方法——属性均值聚类属性支持向量机(AMC-ASVM)

孙喜晨1,贺仁亚2,封举富1   

Abstract: A novel pattern classification method, named AMC-ASVM, was proposed based on Attribute Mean Clustering and Support Vector Machine. The advantages of AMC-ASVM include robustness and high efficiency. AMC ASVM is the natural extension of Cluster Nearest Neighbor Classification as well. Application in Microarray Analysis shows the performance of the present method.

Key words: pattern recognition, attribute mean clustering, support vector machine, gene exp ression data

摘要: 在属性均值聚类(AMC)与支持向量机(SVM)的基础上,提出了一个新的模式分类算法——基于(属性)聚类的属性支持向量机算法(AMC-ASVM)。主要思想是利用属性均值聚类网络得到的具有概率信息(权重)的样本,来训练属性支持向量机,从而得到分类器。这种方法结合了属性聚类的稳定性与属性支持向量机可以利用加权样本的优点,适合处理具有强噪声的数据。另外,该方法也可以看作是堆近邻分类法的自然推广。在实验部分,将其用于结肠癌基因表达数据的处理。实验结果显示了AMC-ASVM在一定程度上优于最近邻, Boosting, 堆近邻, SVM等方法。

关键词: 模式识别, 属性均值聚类, 支持向量机, 基因表达数据

CLC Number: