A Kernel Multiple Change-point Algorithm ...
Document type :
Article dans une revue scientifique
Permalink :
Title :
A Kernel Multiple Change-point Algorithm via Model Selection
Author(s) :
Arlot, Sylvain [Auteur]
Celisse, Alain [Auteur]
Laboratoire Paul Painlevé - UMR 8524
Laboratoire Paul Painlevé - UMR 8524 [LPP]
Harchaoui, Zaid [Auteur]
Celisse, Alain [Auteur]

Laboratoire Paul Painlevé - UMR 8524
Laboratoire Paul Painlevé - UMR 8524 [LPP]
Harchaoui, Zaid [Auteur]
Journal title :
Journal of Machine Learning Research
Abbreviated title :
JMLR
Volume number :
20
Pages :
1-56
Publisher :
Microtome Publishing
Publication date :
2019-12
ISSN :
1532-4435
Keyword(s) :
Change-point detection
Kernel methods
Model selection
Concentration inequality
Concentration in- equality
Kernel methods
Model selection
Concentration inequality
Concentration in- equality
HAL domain(s) :
Mathématiques [math]/Statistiques [math.ST]
English abstract : [en]
We tackle the change-point problem with data belonging to a general set. We build a penalty for choosing the number of change-points in the kernel-based method of Harchaoui and Cappé (2007). This penalty generalizes the ...
Show more >We tackle the change-point problem with data belonging to a general set. We build 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 by Lebarbier (2005) for one-dimensional signals. We prove a non-asymptotic oracle inequality for the proposed method, thanks to a new concentration result for some function of Hilbert-space valued random variables. Experiments on synthetic data illustrate the accuracy of our method, showing that it can detect changes in the whole distribution of data, even when the mean and variance are constant.Show less >
Show more >We tackle the change-point problem with data belonging to a general set. We build 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 by Lebarbier (2005) for one-dimensional signals. We prove a non-asymptotic oracle inequality for the proposed method, thanks to a new concentration result for some function of Hilbert-space valued random variables. Experiments on synthetic data illustrate the accuracy of our method, showing that it can detect changes in the whole distribution of data, even when the mean and variance are constant.Show less >
Language :
Anglais
Audience :
Internationale
Popular science :
Non
Administrative institution(s) :
CNRS
Université de Lille
Université de Lille
Submission date :
2020-06-08T14:10:41Z
2020-06-09T09:03:47Z
2020-06-09T09:03:47Z
Files
- documen
- Open access
- Access the document