Presentation 2007-12-22
Probabilistic Outputs for Multiclass Support Vector Machines
Liu LIU, Mauricio KUGLER, Susumu KUROYANAGI, Akira IWATA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Support Vector Machines (SVM) have been successfully applied in many classification tasks with great generalization performance. However, the output function of SVMs gives an uncalibrated value, impairing the post-processing and making the combination of several classifiers inefficient, as in the case of multiclass SVMs. Some methods of transforming the binary SVM output in a calibrated posterior probability have been proposed, notably the sigmoid fitting method by Platt. This paper proposes an extension of the Platt's model for multiclass SVMs, by combining the optimization procedures of all sigmoid functions. Experimental results are presented and confirm the efficiency of the proposed method.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) support vector machines / multiclass classification / posterior probability / sigmoid fitting / optimization procedure
Paper # NC2007-74
Date of Issue

Conference Information
Committee NC
Conference Date 2007/12/15(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Neurocomputing (NC)
Language ENG
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Probabilistic Outputs for Multiclass Support Vector Machines
Sub Title (in English)
Keyword(1) support vector machines
Keyword(2) multiclass classification
Keyword(3) posterior probability
Keyword(4) sigmoid fitting
Keyword(5) optimization procedure
1st Author's Name Liu LIU
1st Author's Affiliation The authors are with the Department of Computer Science & Engineering, Nagoya Institute of Technology()
2nd Author's Name Mauricio KUGLER
2nd Author's Affiliation The authors are with the Department of Computer Science & Engineering, Nagoya Institute of Technology
3rd Author's Name Susumu KUROYANAGI
3rd Author's Affiliation The authors are with the Department of Computer Science & Engineering, Nagoya Institute of Technology
4th Author's Name Akira IWATA
4th Author's Affiliation The authors are with the Department of Computer Science & Engineering, Nagoya Institute of Technology
Date 2007-12-22
Paper # NC2007-74
Volume (vol) vol.107
Number (no) 410
Page pp.pp.-
#Pages 6
Date of Issue