We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

math.OC

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Mathematics > Optimization and Control

Title: A semismooth Newton-proximal method of multipliers for $\ell_1$-regularized convex quadratic programming

Abstract: In this paper we present a method for the solution of $\ell_1$-regularized convex quadratic optimization problems. It is derived by suitably combining a proximal method of multipliers strategy with a semi-smooth Newton method. The resulting linear systems are solved using a Krylov-subspace method, accelerated by appropriate general-purpose preconditioners, which are shown to be optimal with respect to the proximal parameters. Practical efficiency is further improved by warm-starting the algorithm using a proximal alternating direction method of multipliers. We show that the method achieves global convergence under feasibility assumptions. Furthermore, under additional standard assumptions, the method can achieve global linear and local superlinear convergence. The effectiveness of the approach is numerically demonstrated on $L^1$-regularized PDE-constrained optimization problems.
Subjects: Optimization and Control (math.OC); Numerical Analysis (math.NA)
Cite as: arXiv:2201.10211 [math.OC]
  (or arXiv:2201.10211v1 [math.OC] for this version)

Submission history

From: Spyridon Pougkakiotis [view email]
[v1] Tue, 25 Jan 2022 10:07:44 GMT (42kb)

Link back to: arXiv, form interface, contact.