Open Access
2018 A quasi-Bayesian perspective to online clustering
Le Li, Benjamin Guedj, Sébastien Loustau
Electron. J. Statist. 12(2): 3071-3113 (2018). DOI: 10.1214/18-EJS1479


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 for which we give a convergence guarantee. Finally, numerical experiments illustrate the potential of our procedure.


Download Citation

Le Li. Benjamin Guedj. Sébastien Loustau. "A quasi-Bayesian perspective to online clustering." Electron. J. Statist. 12 (2) 3071 - 3113, 2018.


Received: 1 December 2017; Published: 2018
First available in Project Euclid: 20 September 2018

zbMATH: 06942966
MathSciNet: MR3856169
Digital Object Identifier: 10.1214/18-EJS1479

Primary: 62L12
Secondary: 62C10 , 62C20 , 62L20

Keywords: minimax regret bounds , Online clustering , quasi-Bayesian learning , reversible jump Markov chain Monte Carlo

Vol.12 • No. 2 • 2018
Back to Top