SVM-Maj: a majorization approach to linear support vector machines with different hinge errors

Patrick Groenen (), G.. Nalbantov and J.C. Bioch
Additional contact information
J.C. Bioch: Erasmus Econometric Institute

No EI 2007-49 Revision_Date: 2009-11-06, Econometric Institute Report from Erasmus University Rotterdam, Econometric Institute

Abstract: Support vector machines (SVM) are becoming increasingly popular for the prediction of a binary dependent variable. SVMs perform very well with respect to competing techniques. Often, the solution of an SVM is obtained by switching to the dual. In this paper, we stick to the primal support vector machine (SVM) problem, study its effective aspects, and propose varieties of convex loss functions such as the standard for SVM with the absolute hinge error as well as the quadratic hinge and the Huber hinge errors. We present an iterative majorization algorithm that minimizes each of the adaptations. In addition, we show that many of the features of an SVM are also obtained by an optimal scaling approach to regression. We illustrate this with an example from the literature and do a comparison of different methods on several empirical data sets.

Keywords: support vector machines; iterative majorization; I-splines; absolute hinge error; quadratic hinge error; huber hinge error; optimal scaling (search for similar items in EconPapers)
Date: 2007-11-01

Downloads: (external link)
http://hdl.handle.net/1765/12011 (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: http://EconPapers.repec.org/RePEc:dgr:eureir:1765012011

Access Statistics for this paper

More papers in Econometric Institute Report from Erasmus University Rotterdam, Econometric Institute
Series data maintained by Anneke Kop ().