Titre :

A Quasi-Bayesian Perspective to Online Clustering

Auteur(s) :

Li, Le [Auteur]
Guedj, Benjamin [Auteur]
Laboratoire Paul Painlevé - UMR 8524 [LPP]
Loustau, Sébastien [Auteur]

Titre de la revue :

Electronic journal of statistics

Nom court de la revue :

Electron. j. stat.

Éditeur :

Shaker Heights, OH : Institute of Mathematical Statistics

Date de publication :

2018

ISSN :

1935-7524

Mot(s)-clé(s) :

Online clustering
Minimax regret bounds
Reversible Jump Markov Chain Monte Carlo
Quasi-Bayesian learning

Discipline(s) HAL :

Statistiques [stat]/Machine Learning [stat.ML]

Résumé en anglais : [en]

When faced with high frequency streams of data, clustering raises theoretical and algorithmic pitfalls. We introduce a new and adaptive online clustering algorithm relying on a quasi-Bayesian approach, with a dynamic (i.e., ...
Lire la suite >When faced with high frequency streams of data, clustering raises theoretical and algorithmic pitfalls. We introduce a new and adaptive online clustering algorithm relying on a quasi-Bayesian approach, with a dynamic (i.e., time-dependent) estimation of the (unknown and changing) number of clusters. We prove that our approach is supported by minimax regret bounds. We also provide an RJMCMC-flavored implementation (called PACBO, see https:\/\/cran.r-project.org\/web\/packages\/PACBO\/index.html) for which we give a convergence guarantee. Finally, numerical experiments illustrate the potential of our procedure.Lire moins >

Langue :

Anglais

Audience :

Internationale

Vulgarisation :

Non

Établissement(s) :

CNRS
Université de Lille

Collections :

• METRICS : Evaluation des technologies de santé et des pratiques médicales - ULR 2694

Date de dépôt :

2020-06-08T14:10:27Z
2020-06-09T08:55:33Z