hal-00373805, version 1
Kernel Projection Machine: selection de modeles pour ce nouvel algorithme de classification
Laurent Zwald
1Gilles Blanchard 2
37 eme journee de statistiques sfds (2005)
Abstract: This work presents a theoretical study of the Kernel Projection Machine, a classification algorithm which is an alternate to the Support Vector Machine (SVM). We propose to replace the Tikhonov regularization in the SVM by a penalized finite-dimensional projection. We show that a penalty term proportional to the dimension is appropriate under a ``gap'' hypothesis on the ditribution of class given observation. We also show how to apply this algorithm in practice and present examples of numerical results.
- 1: Laboratoire Jean Kuntzmann (LJK)
- CNRS : UMR5224 – Université Joseph Fourier - Grenoble I – Université Pierre Mendès-France - Grenoble II – Institut Polytechnique de Grenoble - Grenoble Institute of Technology
- 2: Fraunhofer First (IDA)
- Fraunhofer FIRST
- Domain : Statistics/Machine Learning
- hal-00373805, version 1
- http://hal.archives-ouvertes.fr/hal-00373805
- oai:hal.archives-ouvertes.fr:hal-00373805
- From: Laurent Zwald
- Submitted on: Tuesday, 7 April 2009 14:13:58
- Updated on: Tuesday, 7 April 2009 14:36:54
Associated documents
Export