最新Prototype Classification.pdfVIP

Prototype Classification: Insights from Machine Learning ∗ † Arnulf B.A. Graf, Olivier Bousquet, ‡ § Gunnar R¨atsch Bernhard Sch¨olkopf Max Planck Institute for Biological Cybernetics, T¨ubingen, Germany March 21, 2008 Abstract We shed light on the discrimination between patterns belonging to two different classes by casting this decoding problem into a generalized prototype framework. The discrimination process is then separated into two stages: a projection stage that reduces the dimensionality of the data by projecting it on a line, and a threshold stage where the distributions of the projected patterns of both classes are separated. For this, we extend the popular mean-of-class prototype classification using algo- rithms from machine learning that satisfy a set of invariance properties. We report a simple, yet general, approach to express different types of linear classification al- gorithms in an identical and easy-to-visualize formal framework using generalized prototypes where these prototypes are used to express the normal vector and offset of the hyperplane. We investigate non-margin classifiers such as the classical proto- type classifier, the Fisher classifier and the relevance vector machine. We then study hard and soft margin classifiers such as the support vector machine, and a boosted ∗Present address: Center for Neural Science, New York University, USA – email: arnulf.graf@ †Present address: Pertinence, France – email: olivier.bousquet@ ‡Present address: Friedrich Miescher Laboratory of the Max Planck Society, Germany – email: Gunnar.Raetsch@tuebingen.mpg.de §email: ber