# Hybrid and Iteratively Reweighted Regularization by Unbiased Predictive Risk and Weighted GCV for Projected Systems

@article{Renaut2017HybridAI, title={Hybrid and Iteratively Reweighted Regularization by Unbiased Predictive Risk and Weighted GCV for Projected Systems}, author={Rosemary Anne Renaut and Saeed Vatankhah and Vahid Ebrahimzadeh Ardestani}, journal={SIAM J. Sci. Comput.}, year={2017}, volume={39} }

Tikhonov regularization for projected solutions of large-scale ill-posed problems is considered. The Golub-Kahan iterative bidiagonalization is used to project the problem onto a subspace and regularization then applied to find a subspace approximation to the full problem. Determination of the regularization parameter using the method of unbiased predictive risk estimation is considered and contrasted with the generalized cross validation and discrepancy principle techniques. Examining the… Expand

#### Figures, Tables, and Topics from this paper

#### 35 Citations

Unbiased predictive risk estimation of the Tikhonov regularization parameter: convergence with increasing rank approximations of the singular value decomposition

- Mathematics
- BIT Numerical Mathematics
- 2019

The truncated singular value decomposition may be used to find the solution of linear discrete ill-posed problems in conjunction with Tikhonov regularization and requires the estimation of a… Expand

Large-scale Inversion of Magnetic Data Using Golub-Kahan Bidiagonalization with Truncated Generalized Cross Validation for Regularization Parameter Estimation

- Mathematics
- 2018

In this paper a fast method for large-scale sparse inversion of magnetic data is considered. The L1-norm stabilizer is used to generate models with sharp and distinct interfaces. To deal with the… Expand

Convergence of Regularization Parameters for Solutions Using the Filtered Truncated Singular Value Decomposition

- Mathematics
- 2018

The truncated singular value decomposition (TSVD) may be used to find the solution of the linear discrete ill-posed problem Ax ≈ b using Tikhonov regularization. Regularization parameter α balances… Expand

Fast and sparse inversion of magnetic data with truncated generalized cross validation

- 2016

In this paper a fast method for sparse inversion of magnetic data is considered. The L1-norm stabilizer is used to generate models with sharp and distinct interfaces. To deal with the non-linearity… Expand

Generalized Hybrid Iterative Methods for Large-Scale Bayesian Inverse Problems

- Mathematics, Computer Science
- SIAM J. Sci. Comput.
- 2017

A hybrid algorithm based on the generalized Golub-Kahan bidiagonalization for computing Tikhonov regularized solutions to problems where explicit computation of the square root and inverse of the covariance kernel for the prior covariance matrix is not feasible. Expand

Adaptive Regularization Parameter Choice Rules for Large-Scale Problems

- Computer Science, Mathematics
- ArXiv
- 2019

This paper derives a new class of adaptive regularization parameter choice strategies that can be effectively and efficiently applied when regularizing large-scale linear inverse problems by… Expand

Hybrid projection methods for large-scale inverse problems with mixed Gaussian priors

- Physics, Mathematics
- Inverse Problems
- 2020

When solving ill-posed inverse problems, a good choice of the prior is critical for the computation of a reasonable solution. A common approach is to include a Gaussian prior, which is defined by a… Expand

3-D Projected $ L_{1}$ inversion of gravity data

- Mathematics, Physics
- 2016

Sparse inversion of gravity data based on $L_1$-norm regularization is discussed. An iteratively reweighted least squares algorithm is used to solve the problem. At each iteration the solution of a… Expand

A fast algorithm for regularized focused 3-D inversion of gravity data using the randomized SVD

- Mathematics
- 2017

A fast algorithm for solving the under-determined 3-D linear gravity inverse problem based on the randomized singular value decomposition (RSVD) is developed. The algorithm combines an iteratively… Expand

Krylov Methods for Low-Rank Regularization

- Computer Science, Mathematics
- SIAM J. Matrix Anal. Appl.
- 2020

New solvers for the computation of low-rank approximate solutions to large-scale linear problems, with a particular focus on the regularization of linear inverse problems are introduced. Expand

#### References

SHOWING 1-10 OF 63 REFERENCES

Regularization parameter estimation for underdetermined problems by the χ 2 principle with application to 2D focusing gravity inversion

- Mathematics
- 2014

The principle generalizes the Morozov discrepancy principle to the augmented residual of the Tikhonov regularized least squares problem. For weighting of the data fidelity by a known Gaussian noise… Expand

Application of the χ2 principle and unbiased predictive risk estimator for determining the regularization parameter in 3-D focusing gravity inversion

- Mathematics
- 2015

The χ 2 principle and the unbiased predictive risk estimator are used to determine optimal regularization parameters in the context of 3D focusing gravity inversion with the minimum support… Expand

Regularization parameter estimation for large-scale Tikhonov regularization using a priori information

- Mathematics, Computer Science
- Comput. Stat. Data Anal.
- 2010

An approach using the Golub-Kahan iterative bidiagonalization of the regularized problem is presented and is concluded that the presented approach is robust for both small and large-scale discretely ill-posed least squares problems. Expand

Some results on the regularization of LSQR for large-scale discrete ill-posed problems

- Mathematics
- 2015

LSQR, a Lanczos bidiagonalization based Krylov subspace iterative method, and its mathematically equivalent conjugate gradient for least squares problems (CGLS) applied to normal equations system,… Expand

The regularizing effect of the Golub-Kahan iterative bidiagonalization and revealing the noise level in the data

- Mathematics
- 2009

Regularization techniques based on the Golub-Kahan iterative bidiagonalization belong among popular approaches for solving large ill-posed problems. First, the original problem is projected onto a… Expand

A weighted-GCV method for Lanczos-hybrid regularization.

- Mathematics
- 2007

Lanczos-hybrid regularization methods have been proposed as effective approaches for solving largescale ill-posed inverse problems. Lanczos methods restrict the solution to lie in a Krylov subspace,… Expand

Generalized cross-validation as a method for choosing a good ridge parameter

- Mathematics, Computer Science
- Milestones in Matrix Computation
- 2007

The method of generalized cross-validation (GCV) for choosing a good value for λ from the data is studied, which can be used in subset selection and singular value truncation methods for regression, and even to choose from among mixtures of these methods. Expand

Rank-Deficient and Discrete Ill-Posed Problems: Numerical Aspects of Linear Inversion

- Mathematics
- 1987

Preface Symbols and Acronyms 1. Setting the Stage. Problems With Ill-Conditioned Matrices Ill-Posed and Inverse Problems Prelude to Regularization Four Test Problems 2. Decompositions and Other… Expand

Windowed Spectral Regularization of Inverse Problems

- Mathematics, Computer Science
- SIAM J. Sci. Comput.
- 2011

This work shows how to use standard parameter-choice methods, such as the discrepancy principle and generalized cross-validation, in a windowed regularization framework, and demonstrates the effectiveness of the algorithms on deblurring images and on the backward heat equation. Expand

Choosing Regularization Parameters in Iterative Methods for Ill-Posed Problems

- Mathematics, Computer Science
- SIAM J. Matrix Anal. Appl.
- 2001

This work presents a common framework for efficient algorithms that regularize after this second projection rather than before it, and shows that determining regularization parameters based on the final projected problem rather than on the original discretization has firmer justification and often involves less computational expense. Expand