Skip to Main content Skip to Navigation
Conference papers

A diffusion strategy for distributed dictionary learning

Pierre Chainais 1, 2, 3, 4 Cédric Richard 5
LAGIS - Laboratoire d'Automatique, Génie Informatique et Signal
4 SEQUEL - Sequential Learning
LIFL - Laboratoire d'Informatique Fondamentale de Lille, Inria Lille - Nord Europe, LAGIS - Laboratoire d'Automatique, Génie Informatique et Signal
Abstract : We consider the problem of a set of nodes which is required to collectively learn a common dictionary from noisy measurements. This distributed dictionary learning approach may be useful in several contexts including sensor networks. Dif-fusion cooperation schemes have been proposed to estimate a consensus solution to distributed linear regression. This work proposes a diffusion-based adaptive dictionary learning strategy. Each node receives measurements which may be shared or not with its neighbors. All nodes cooperate with their neighbors by sharing their local dictionary to estimate a common representa-tion. In a diffusion approach, the resulting algorithm corresponds to a distributed alternate optimization. Beyond dictionary learn-ing, this strategy could be adapted to many matrix factorization problems in various settings. We illustrate its efficiency on some numerical experiments, including the difficult problem of blind hyperspectral images unmixing.
Complete list of metadata

Cited literature [8 references]  Display  Hide  Download
Contributor : Pierre Chainais Connect in order to contact the contributor
Submitted on : Monday, January 19, 2015 - 11:20:42 AM
Last modification on : Tuesday, October 19, 2021 - 7:00:54 PM
Long-term archiving on: : Monday, April 20, 2015 - 10:36:33 AM


Files produced by the author(s)


  • HAL Id : hal-01104781, version 1
  • ARXIV : 1410.0719


Pierre Chainais, Cédric Richard. A diffusion strategy for distributed dictionary learning. 2nd "international Traveling Workshop on Interactions between Sparse models and Technology" (iTWIST'14), Laurent Jacques, Aug 2014, Namur, Belgium. ⟨hal-01104781⟩



Record views


Files downloads