Computer Engineering and Applications ›› 2007, Vol. 43 ›› Issue (10): 76-78.

• 学术探讨 • Previous Articles     Next Articles

A General Formulation of 1st-Order Polynomial Smooth Support Vector Machines for Classification

jinzhi xiong   

  • Received:2006-10-05 Revised:1900-01-01 Online:2007-04-01 Published:2007-04-01
  • Contact: jinzhi xiong

一阶多项式光滑的支持向量分类机的一般模型

熊金志 李广明 高晓雷 牛熠   

  1. 东莞理工学院软件学院 东莞理工学院计算中心 华南理工大学应用数学系
  • 通讯作者: 熊金志

Abstract: This paper derived a class of 1st-order-smooth polynomial functions that are able to approximate the plus function over a common interval around the origin. With such functions, a general model, 1st-Order Smooth Support Vector Machine (1SSVM) was developed. Theoretical analysis shows that the 1st-order smooth function used in [2] belongs to this class and the resultant smooth SVM is also a special case of 1SSVM. Hence this paper theoretically solved the general formulation of 1st-order smooth SVMs

Key words: classification, support vector machine, data mining, smoothing.

摘要: 本文在一个包含原点的一般区间导出了一类光滑正号函数的一阶多项式函数,还研究了用此类函数对支持向量机作光滑处理的问题,提出了一阶光滑的支持向量机的一般模型 1SSVM(1st—Order Smooth Support Vector Machine)。理论分析表明,文献[2]中所用的一阶光滑函数是此类函数的一个特例,其提出的一阶光滑的支持向量机也是模型 1SSVM的一个特例,从而从理论上解决了一阶光滑支持向量机的一般模型问题。

关键词: 分类, 支持向量机, 数据挖掘, 光滑