In high dimensional regression settings, sparsity enforcing penalties have proved useful to regularize the data-fitting term. A recently introduced technique called screening rules propose to ignore some variables in the optimization leveraging the expected sparsity of the solutions and consequently leading to faster solvers. In this work, we propose a unifying framework for generalized linear models regularized with standard sparsity enforcing penalties such as l1 or l1/l2 norms.

This presentation is part of Minisymposium “MS54 - Hybrid Approaches that Combine Deterministic and Statistical Regularization for Applied Inverse Problems (4 parts)”
organized by: Cristiana Sebu (University of Malta) , Taufiquar Khan (Clemson University) .

Authors:

Ndiaye Eugene (Telecom ParisTech)

Olivier Fercoq (Telecom ParisTech)

Alexandre Gramfort (University Paris Saclay)

Joseph Salmon (Telecom ParisTech)

Keywords: