Loading...
Loading...

Go to the content (press return)

A study on output normalization in multiclass SVMs

Author
Gonzalez-Abril, L.; Velasco, F.; Angulo, C.; Ortega, J.
Type of activity
Journal article
Journal
Pattern recognition letters
Date of publication
2013-02-01
Volume
34
Number
3
First page
344
Last page
348
DOI
https://doi.org/10.1016/j.patrec.2012.11.003 Open in new window
Project funding
Saving Energy by Intelligent Monitoring
Repository
http://hdl.handle.net/2117/19156 Open in new window
URL
http://www.sciencedirect.com/science/article/pii/S0167865512003637 Open in new window
Abstract
The use of binary support vector machines (SVMs) in multi-classification is addressed in this paper. Margins associated to the bi-classifiers, since they depend on the geometrical disposition of the classes being separated, are, in general, of various magnitudes. In order to overcome this scaling problem, a normalization process should be applied on the SVMs’ outputs. Thus, a new normalization approach is presented based on the convex hulls that contain the classes to be separated. Furthermore...
Citation
González, L. [et al.]. A study on output normalization in multiclass SVMs. "Pattern recognition letters", 01 Febrer 2013, vol. 34, núm. 3, p. 344-348.
Keywords
1-v-r SVM, Convex hull, Kernel methods, Multiclassification
Group of research
GREC - Knowledge Engineering Research Group
IDEAI-UPC - Intelligent Data Science and Artificial Intelligence Research Center

Participants

  • Gonzalez Abril, Luis  (author)
  • Velasco Morente, Francisco  (author)
  • Angulo Bahón, Cecilio  (author)
  • Ortega Ramirez, Juan Antonio  (author)