Evolving class for SVM's incremental learning.

Ahmed Benzerrouk (1) , Brigitte Chebel-Morello (1) , Noureddine Zerhouni (1)

1 LAB - Laboratoire d'automatique de Besançon

The good generalization performance of support vector machines (SVM) has made them a popular tool in artificial intelligence community. In this paper, we prove that SVM multi class algorithms are not equivalent for all classification problems we present a new approach for incremental learning using SVM that create a rejection class which would be interesting for fault diagnosis where fault classes usually evolve with time : It is when some new samples may be rejected by all the current classes. Hence, these samples may correspond to a new fault (a new class) which may appear after the first training step.

Domaines: Automatique / Robotique

Ahmed Benzerrouk, Brigitte Chebel-Morello, Noureddine Zerhouni. Evolving class for SVM's incremental learning.. 16th International Conference on Computing in section Artificial Intelligence and Applications, ICC'07., Nov 2007, Mexico City, Mexico. 6 p. ⟨hal-00342432⟩