Tabu search model selection for SVM

Gilles Lebrun (1) , Christophe Charrier (2) , Olivier Lezoray (2) , Hubert Cardot (3)

1 GREYC - Groupe de Recherche en Informatique, Image et Instrumentation de Caen
2 Equipe Image - Laboratoire GREYC - UMR6072
3 LIFAT - Laboratoire d'Informatique Fondamentale et Appliquée de Tours

A model selection method based on tabu search is proposed to build support vector machines (binary decision functions) of reduced complexity and efficient generalization. The aim is to build a fast and efficient support vector machines classifier. A criterion is defined to evaluate the decision function quality which blends recognition rate and the complexity of a binary decision functions together. The selection of the simplification level by vector quantization, of a feature subset and of support vector machines hyperparameters are performed by tabu search method to optimize the defined decision function quality criterion in order to find a good sub-optimal model on tractable times.

Domaines: Traitement des images [eess.IV]; Réseau de neurones [cs.NE]

Gilles Lebrun, Christophe Charrier, Olivier Lezoray, Hubert Cardot. Tabu search model selection for SVM. International Journal of Neural Systems, 2008, 18 (1), pp.19-31. ⟨hal-00330025⟩