计算机科学与探索 ›› 2013, Vol. 7 ›› Issue (8): 754-761.DOI: 10.3778/j.issn.1673-9418.1206054

• 学术研究 • 上一篇    下一篇

粒计算与统计学习理论

张  铃1,钱付兰1+,何富贵2   

  1. 1. 安徽大学 计算机科学与技术学院,合肥 230039
    2. 皖西学院 信息工程学院,安徽 六安 237012
  • 出版日期:2013-08-01 发布日期:2013-08-06

Granular Computing and Statistical Learning

ZHANG Ling1, QIAN Fulan1+, HE Fugui2   

  1. 1. School of Computer Science and Technology, Anhui University, Hefei 230039, China
    2. Department of Information Engineering, West Anhui University, Lu’an, Anhui 237012, China
  • Online:2013-08-01 Published:2013-08-06

摘要: 互联网信息时代,如何从复杂的数据中进行有目标的数据挖掘是很多领域的一个中心问题。目前针对此问题的方法大多是基于统计学习理论的机器学习方法,并且粒计算在数据挖掘问题中有着广泛的应用。将粒计算方法与统计学习方法相结合,提出了一个更优的粒计算统计学习方法。给出了一个基于粒计算的统计分类算法,并与支持向量机(support vector machine,SVM)、覆盖算法进行了比较,实验表明通过粒化所得到的支持向量求解出的分类结果较优。

关键词: 统计学习理论, 粒计算, 商空间, 支持向量机(SVM)

Abstract: In Internet information age, how to operate objective data mining based on complex date remains a central problem in many fields. The major approach to the problem is machine learning based on statistical learning theory. Granular computing has a wide range of applications in data mining. This paper proposes a method to get a better granular computing and statistical learning method by combining statistical learning theory and granular computing. Finally, this paper gives a statistical classification algorithm based on granular computing, and compares the results with support vector machine (SVM) and covering algorithm. The experimental results show that the classification effect of granulized support vector is better.

Key words: statistical learning theory, granular computing, quotient space, support vector machine (SVM)