A greedy algorithm for optimizing the kernel alignment and the performance of kernel machines

Jean-Baptiste Pothin 1 Cédric Richard 1
1 M2S - Modélisation et Sûreté des Systèmes
ISTIT - Institut des Sciences et Technologies de l'Information de Troyes
Abstract : Kernel-target alignment has recently been proposed as a criterion for measuring the degree of agreement between a reproducing kernel and a learning task. It makes possible to find a powerful kernel for a given classification problem without designing any classifier. In this paper, we present an alternating optimization strategy, based on a greedy algorithm for maximizing the alignment over linear combinations of kernels, and a gradient descent to adjust the free parameters of each kernel. Experimental results show an improvement in the classification performance of support vector machines, and a drastic reduction in the training time.
Document type :
Conference papers
Complete list of metadatas

https://hal-utt.archives-ouvertes.fr/hal-02284284
Contributor : Jean-Baptiste Vu Van <>
Submitted on : Wednesday, September 11, 2019 - 4:10:26 PM
Last modification on : Monday, September 16, 2019 - 4:36:04 PM

Identifiers

  • HAL Id : hal-02284284, version 1

Collections

UTT

Citation

Jean-Baptiste Pothin, Cédric Richard. A greedy algorithm for optimizing the kernel alignment and the performance of kernel machines. 2006 14th European Signal Processing Conference, Sep 2006, Florence, Italy. ⟨hal-02284284⟩

Share

Metrics

Record views

4