hal-00671174, version 1
Kernel change-point detection
(2012-02-14)
Résumé : We tackle the change-point problem with data belonging to a general set. We propose a penalty for choosing the number of change-points in the kernel-based method of Harchaoui and Cappé (2007). This penalty generalizes the one proposed for one dimensional signals by Lebarbier (2005). We prove it satisfies a non-asymptotic oracle inequality by showing a new concentration result in Hilbert spaces. Experiments on synthetic and real data illustrate the accuracy of our method, showing it can detect changes in the whole distribution of data, even when the mean and variance are constant. Our algorithm can also deal with data of complex nature, such as the GIST descriptors which are commonly used for video temporal segmentation.
- 1 :
- CNRS : UMR8548 – Ecole normale supérieure de Paris - ENS Paris
- 2 :
- INRIA : PARIS - ROCQUENCOURT – Ecole normale supérieure de Paris - ENS Paris – CNRS : UMR8548
- 3 :
- CNRS : UMR8524 – Université Lille I - Sciences et technologies
- 4 :
- INRIA
- 5 :
- CNRS : UMR5527 – INRIA – Laboratoire Jean Kuntzmann – Université Joseph Fourier - Grenoble I – Institut National Polytechnique de Grenoble (INPG)
- Collaboration : ANR Detect
- Domaine : Mathématiques/Statistiques
Statistiques/Théorie - Mots-clés : model selection – kernel methods – change-point problem – concentration inequality
- hal-00671174, version 1
- http://hal.archives-ouvertes.fr/hal-00671174
- oai:hal.archives-ouvertes.fr:hal-00671174
- Contributeur :
- Soumis le : Vendredi 17 Février 2012, 11:50:59
- Dernière modification le : Vendredi 17 Février 2012, 11:59:59
Documents associés
Exporter