F/sub beta / support vector machines
(2005) Proceedings of the International Joint Conference on Neural Networks 2005 — Location: Montreal, Que., Canada (31.July.2005)
Files
No attached file found for this publication.
Details
- Authors
- Abstract
- We introduce in this paper F/sub beta / SVMs, a new parametrization of support vector machines. It allows to optimize a SVM in terms of F/sub beta /, a classical information retrieval criterion, instead of the usual classification rate. Experiments illustrate the advantages of this approach with respect to the traditional 2-norm soft-margin SVM when precision and recall are of unequal importance. An automatic model selection procedure based on the generalization F/sub beta / score is introduced. It relies on the results of Chapelle, Vapnjk et al. (2002) about the use of gradient-based techniques in SVM model selection. The derivatives of a F /sub beta / loss function with respect to the hyperparameters C and the width sigma of a gaussian kernel are formally defined. The model is then selected by performing a gradient descent of the F/sub beta / loss function over the set of hyperparameters. Experiments on artificial and real-life data show the benefits of this method when the F/sub beta / score is considered.
- Affiliations
Citations
Callut, J., & Dupont, P. (2005). F/sub beta / support vector machines. Proceedings of the International Joint Conference on Neural Networks2005 (IEEE Cat. No. 05CH37663C), Vol. 3, p. 1443-8. https://hdl.handle.net/2078.5/218874