Convergence rates for the relaxed Peaceman-Rachford splitting method on a monotone inclusion problem

Research output: Contribution to journalArticlepeer-review

7 Downloads (Pure)

Abstract

We consider the convergence behavior using the relaxed Peaceman-Rachford splitting method to solve the monotone inclusion problem 0 ∈ (A + B)(u), where A, : R⇒ Rn are maximal β-strongly monotone operators, n ≥ 1 and β > 0. Under a technical assumption, convergence of iterates using the method on the problem is proved when either A or B is single-valued, and the fixed relaxation parameter θ lies in the interval (2 + β, 2 + β + min {β, 1/β}). With this convergence result, we address an open problem that is not settled in [22] on the convergence of these iterates for θ ∈ in (2 + β, 2 + β + min{β, 1/β}). Pointwise convergence rate results and R-linear convergence rate results when θ lies in the interval [2 + β, 2 + β + min {β, 1/β}) are also provided in the paper. Our analysis to achieve these results is atypical and hence novel. Numerical experiments on the weighted Lasso minimization problem are conducted to test the validity of the assumption.
Original languageEnglish
Pages (from-to)298–323
JournalJournal of Optimization Theory and Applications
Volume196
Early online date2 Dec 2022
DOIs
Publication statusPublished - 1 Jan 2023

Keywords

  • Relaxed Peaceman-Rachford splitting method
  • maximal strong monotonicity
  • convergence
  • pointwise convergence rate
  • R-linear convergence rate

Fingerprint

Dive into the research topics of 'Convergence rates for the relaxed Peaceman-Rachford splitting method on a monotone inclusion problem'. Together they form a unique fingerprint.

Cite this