Skip to Main content Skip to Navigation
Journal articles

Proximal Gradient methods with Adaptive Subspace Sampling

Abstract : Many applications in machine learning or signal processing involve nonsmooth optimization problems. This nonsmoothness brings a low-dimensional structure to the optimal solutions. In this paper, we propose a randomized proximal gradient method harnessing this underlying structure. We introduce two key components: i) a random subspace proximal gradient algorithm; ii) an identification-based sampling of the subspaces. Their interplay brings a significant performance improvement on typical learning problems in terms of dimensions explored.
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-02555292
Contributor : Franck Iutzeler <>
Submitted on : Monday, April 27, 2020 - 2:45:52 PM
Last modification on : Monday, May 11, 2020 - 4:35:57 PM

Files

sub_desc.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-02555292, version 1
  • ARXIV : 2004.13356

Collections

Citation

Dmitry Grishchenko, Franck Iutzeler, Jérôme Malick. Proximal Gradient methods with Adaptive Subspace Sampling. Mathematics of Operations Research, INFORMS, In press. ⟨hal-02555292⟩

Share

Metrics

Record views

26

Files downloads

21