Cost-sensitive universum-SVM

Sauptik Dhar, Vladimir Cherkassky

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Scopus citations

Abstract

Many applications of machine learning involve analysis of sparse high-dimensional data, where the number of input features is larger than the number of data samples. Standard classification methods may not be sufficient for such data, and this provides motivation for non-standard learning settings. One such new learning methodology is called Learning through Contradictions or Universum support vector machine (U-SVM) [1, 2]. Recent studies [2-10] have shown U-SVM to be quite effective for such sparse high-dimensional data settings. However, these studies use balanced data sets with equal misclassification costs. This paper extends the U-SVM for problems with different misclassification costs, and presents practical conditions for the effectiveness of the cost sensitive U-SVM. Finally, several empirical comparisons are presented to illustrate the utility of the proposed approach.

Original languageEnglish (US)
Title of host publicationProceedings - 2012 11th International Conference on Machine Learning and Applications, ICMLA 2012
Pages220-225
Number of pages6
DOIs
StatePublished - Dec 1 2012
Event11th IEEE International Conference on Machine Learning and Applications, ICMLA 2012 - Boca Raton, FL, United States
Duration: Dec 12 2012Dec 15 2012

Publication series

NameProceedings - 2012 11th International Conference on Machine Learning and Applications, ICMLA 2012
Volume1

Other

Other11th IEEE International Conference on Machine Learning and Applications, ICMLA 2012
Country/TerritoryUnited States
CityBoca Raton, FL
Period12/12/1212/15/12

Keywords

  • Cost-sensitive SVM
  • Universum SVM
  • learning through contradiction
  • support vector machines

Fingerprint

Dive into the research topics of 'Cost-sensitive universum-SVM'. Together they form a unique fingerprint.

Cite this