A probabilistic incremental proximal gradient method

Abstract : In this paper, we propose a probabilistic optimization method, named probabilistic incremental proximal gradient (PIPG) method, by developing a probabilistic interpretation of the incremental proximal gradient algorithm. We explicitly model the update rules of the incremental proximal gradient method and develop a systematic approach to propagate the uncertainty of the solution estimate over iterations. The PIPG algorithm takes the form of Bayesian filtering updates for a state-space model constructed by using the cost function. Our framework makes it possible to utilize well-known exact or approximate Bayesian filters, such as Kalman or extended Kalman filters, to solve large-scale regularized optimization problems.
Document type :
Preprints, Working Papers, ...
Liste complète des métadonnées

Contributor : Emilie Chouzenoux <>
Submitted on : Thursday, December 6, 2018 - 11:27:19 AM
Last modification on : Friday, April 19, 2019 - 4:55:25 PM

Links full text


  • HAL Id : hal-01946642, version 1
  • ARXIV : 1812.01655


Ömer Deniz Akyildiz, Emilie Chouzenoux, Víctor Elvira, Joaquín Míguez. A probabilistic incremental proximal gradient method. 2018. ⟨hal-01946642⟩



Record views