A note on the minimization of a Tikhonov functional with ${\ell }^1$-penalty

Abstract

In this paper, we consider the minimization of a Tikhonov functional with an ${\ell }_1$ penalty for solving linear inverse problems with sparsity constraints. One of the many approaches used to solve this problem uses the Nemskii operator to transform the Tikhonov functional into one with an ${\ell }_2$ penalty term but a nonlinear operator. The transformed problem can then be analyzed and minimized using standard methods. However, by the nature of this transform, the resulting functional is only once continuously differentiable, which prohibits the use of second order methods. Hence, in this paper, we propose a different transformation, which leads to a twice differentiable functional that can now be minimized using efficient second order methods like Newton's method. We provide a convergence analysis of our proposed scheme, as well as a number of numerical results showing the usefulness of our proposed approach.

arXiv comment20 pages, 4 figures

References in corpus (5)

• The rate of convergence of Nesterov's accelerated forward-backward method is actually faster than $1/k^2$
• Linear convergence of iterative soft-thresholding
• A Semismooth Newton Method for Tikhonov Functionals with Sparsity Constraints
• Discretization of variational regularization in Banach spaces
• Image reconstruction with imperfect forward models and applications in deblurring