Abstract:We exploit the similarities between Tikhonov regularization and Bayesian hierarchical models to propose a regularization scheme that acts like a distributed Tikhonov regularization where the amount of regularization varies from component to component. In the standard formulation, Tikhonov regularization compensates for the inherent ill-conditioning of linear inverse problems by augmenting the data fidelity term measuring the mismatch between the data and the model output with a scaled penalty functional. The selection of the scaling is the core problem in Tikhonov regularization. If an estimate of the amount of noise in the data is available, a popular way is to use the Morozov discrepancy principle, stating that the scaling parameter should be chosen so as to guarantee that the norm of the data fitting error is approximately equal to the norm of the noise in the data. A too small value of the regularization parameter would yield a solution that fits to the noise while a too large value would lead to an excessive penalization of the solution. In many applications, it would be preferable to apply distributed regularization, replacing the regularization scalar by a vector valued parameter, allowing different regularization for different components of the unknown, or for groups of them. A distributed Tikhonov-inspired regularization is particularly well suited when the data have significantly different sensitivity to different components, or to promote sparsity of the solution. The numerical scheme that we propose, while exploiting the Bayesian interpretation of the inverse problem and identifying the Tikhonov regularization with the Maximum A Posteriori (MAP) estimation, requires no statistical tools. A combination of numerical linear algebra and optimization tools makes the scheme computationally efficient and suitable for problems where the matrix is not explicitly available.

On the Generalized Discrepancy Principle for Tikhonov Regularization in Hilbert Scales

Discrepancy Curves for Multi-Parameter Regularization

An Analysis of Tikhonov Regularization for Nonlinear Ill-Posed Problems under a General Smoothness Assumption

Multi-parameter Regularization and Its Numerical Realization

Dual Regularized Total Least Squares and Multi-Parameter Regularization

Condition Numbers and Perturbation Analysis for the Tikhonov Regularization of Discrete Ill-Posed Problems.

Multiscale Analysis for Ill-Posed Problems with Semi-Discrete Tikhonov Regularization

Tikhonov Regularization for Stochastic Non-Smooth Convex Optimization in Hilbert Spaces

A discretizing Tikhonov regularization method via modified parameter choice rules

Tikhonov Regularization with Oversmoothing Penalty: Error Bounds in Terms of Distance Functions

Multi-parameter Tikhonov Regularization with the ℓ 0 Sparsity Constraint

Perturbation Identities for Regularized Tikhonov Inverses and Weighted Pseudoinverses

Learning the optimal Tikhonov regularizer for inverse problems

A Joint Bidiagonalization Based Algorithm for Large Scale Linear Discrete Ill-posed Problems in General-Form Regularization

Distributed Tikhonov regularization for ill-posed inverse problems from a Bayesian perspective

Sparse Reconstruction by Means of the Standard Tikhonov Regularization

Inexact Newton regularization methods in Hilbert scales

On saturation of the discrepancy principle for nonlinear Tikhonov regularization in Hilbert spaces

Regularization of linear ill-posed problems involving multiplication operators

Regularized Total Least Squares: Computational Aspects and Error Bounds.

Tikhonov Regularization with L^0-Term Complementing a Convex Penalty: L^1 Convergence under Sparsity Constraints