3-5 juil. 2013 Villeneuve d'Ascq (Lille) (France)

Par auteur > Richard Cédric

Vendredi 5
Apprentissage Automatique

› 16:30 - 17:00 (30min)
Distributed dictionary learning over a sensor network
Pierre Chainais  1, 2@  , Cédric Richard  3@  
1 : INRIA Lille Nord Europe - SequeL  (INRIA Lille)  -  Site web
INRIA, Ecole Centrale de Lille
2 : Laboratoire d'Automatique, Génie Informatique et Signal  (LAGIS)  -  Site web
CNRS : UMR8219, Ecole Centrale de Lille, Université Lille I - Sciences et technologies
LAGIS Ecole Centrale de Lille Cité Scientifique 59655 Villeneuve d'Ascq -  France
3 : Observatoire de la Cote d'Azur  (Laboratoire J.L. LAGRANGE - OCA)  -  Site web
CNRS : UMS7293, Université de Nice Sophia-Antipolis
B.P. 4229 06304 Nice Cedex 4 -  France

We consider the problem of distributed dictionary learning, where a set of nodes is required to collec- tively learn a common dictionary from noisy measure- ments. This approach may be useful in several contexts including sensor networks. Diffusion cooperation schemes have been proposed to solve the distributed linear regression problem. In this work we focus on a diffusion-based adaptive dictionary learning strategy: each node records observations and cooperates with its neighbors by sharing its local dictionary. The resulting algorithm corresponds to a distributed block coordi- nate descent (alternate optimization). Beyond dictionary learning, this strategy could be adapted to many matrix factorization problems and generalized to various settings. This article presents our approach and illustrates its efficiency on some numerical examples.

Personnes connectées : 2