Relaxed regularization for linear inverse problems

Nick Luiken, Tristan Van Leeuwen

Research output: Contribution to journalArticlepeer-review


We consider regularized least-squares problems of the form min {equation presented}. Recently, Zheng et al. [IEEE Access, 7 (2019), pp. 1404-1423], proposed an algorithm called Sparse Relaxed Regularized Regression (SR3) that employs a splitting strategy by introducing an auxiliary variable y and solves min {equation presented}. By minimizing out the variable x, we obtain an equivalent optimization problem min {equation presented}. In our work, we view the SR3 method as a way to approximately solve the regularized problem. We analyze the conditioning of the relaxed problem in general and give an expression for the SVD of Fκ as a function of κ. Furthermore, we relate the Pareto curve of the original problem to the relaxed problem, and we quantify the error incurred by relaxation in terms of κ. Finally, we propose an efficient iterative method for solving the relaxed problem with inexact inner iterations. Numerical examples illustrate the approach.
Original languageEnglish (US)
Pages (from-to)S269-S292
JournalSIAM Journal on Scientific Computing
StatePublished - Jan 1 2020
Externally publishedYes

Bibliographical note

Generated from Scopus record by KAUST IRTS on 2023-10-22

ASJC Scopus subject areas

  • Computational Mathematics
  • Applied Mathematics


Dive into the research topics of 'Relaxed regularization for linear inverse problems'. Together they form a unique fingerprint.

Cite this