The context-tree kernel for strings. - Bio-informatique (CBIO) Accéder directement au contenu
Article Dans Une Revue Neural Networks Année : 2005

The context-tree kernel for strings.

Marco Cuturi
Jean-Philippe Vert

Résumé

We propose a new kernel for strings which borrows ideas and techniques from information theory and data compression. This kernel can be used in combination with any kernel method, in particular Support Vector Machines for string classification, with notable applications in proteomics. By using a Bayesian averaging framework with conjugate priors on a class of Markovian models known as probabilistic suffix trees or context-trees, we compute the value of this kernel in linear time and space while only using the information contained in the spectrum of the considered strings. This is ensured through an adaptation of a compression method known as the context-tree weighting algorithm. Encouraging classification results are reported on a standard protein homology detection experiment, showing that the context-tree kernel performs well with respect to other state-of-the-art methods while using no biological prior knowledge.

Dates et versions

hal-00433583 , version 1 (20-11-2009)

Identifiants

Citer

Marco Cuturi, Jean-Philippe Vert. The context-tree kernel for strings.. Neural Networks, 2005, 18 (8), pp.1111-23. ⟨10.1016/j.neunet.2005.07.010⟩. ⟨hal-00433583⟩
69 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More