Linear convergence of stochastic block-coordinate fixed point algorithms

Abstract : Recent random block-coordinate fixed point algorithms are particularly well suited to large-scale optimization in signal and image processing. These algorithms feature random sweeping rules to select arbitrarily the blocks of variables that are activated over the course of the iterations and they allow for stochastic errors in the evaluation of the operators. The present paper provides new linear convergence results. These convergence rates are compared to those of standard deter-ministic algorithms both theoretically and experimentally in an image recovery problem.
Liste complète des métadonnées

https://hal.archives-ouvertes.fr/hal-01964580
Contributor : Jean-Christophe Pesquet <>
Submitted on : Saturday, December 22, 2018 - 6:01:24 PM
Last modification on : Thursday, February 7, 2019 - 4:34:19 PM
Document(s) archivé(s) le : Saturday, March 23, 2019 - 12:40:46 PM

File

eusipco2018.pdf
Files produced by the author(s)

Identifiers

Citation

Patrick Combettes, Jean-Christophe Pesquet. Linear convergence of stochastic block-coordinate fixed point algorithms. European Signal and Image Processing Conference (EUSIPCO 2018), Sep 2018, Rome, Italy. ⟨10.23919/EUSIPCO.2018.8552941⟩. ⟨hal-01964580⟩

Share

Metrics

Record views

26

Files downloads

29