Parameter Choices for Sparse Regularization with the \ell₁ Norm

DOI10.1088/1361-6420/ACAD22arXiv2205.07170MaRDI QIDQ6399182FDOQ6399182

Authors: Qianru Liu, Rui Wang, Yuesheng Xu, Mingsong Yan

Publication date: 14 May 2022

Abstract: We consider a regularization problem whose objective function consists of a convex fidelity term and a regularization term determined by the

e l l_{1}

norm composed with a linear transform. Empirical results show that the regularization with the

e l l_{1}

norm can promote sparsity of a regularized solution. It is the goal of this paper to understand theoretically the effect of the regularization parameter on the sparsity of the regularized solutions. We establish a characterization of the sparsity under the transform matrix of the solution. When the fidelity term has a special structure and the transform matrix coincides with a identity matrix, the resulting characterization can be taken as a regularization parameter choice strategy with which the regularization problem has a solution having a sparsity of a certain level. We study choices of the regularization parameter so that the regularization term alleviates the ill-posedness and promote sparsity of the resulting regularized solution. Numerical experiments demonstrate that choices of the regularization parameters can balance the sparsity of the solutions of the regularization problem and its approximation to the minimizer of the fidelity function.

Mathematics Subject Classification ID

Linear operators and ill-posed problems, regularization (47A52)

This page was built for publication: Parameter Choices for Sparse Regularization with the $\ell_1$ Norm

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6399182)