Effective new methods for automated parameter selection in regularized inverse problems
From MaRDI portal
Publication:2301416
DOI10.1016/J.APNUM.2020.01.015zbMATH Open1440.65032arXiv1812.11449OpenAlexW3004294979MaRDI QIDQ2301416FDOQ2301416
Authors: Toby Sanders, Rodrigo B. Platte, Robert D. Skeel 
Publication date: 24 February 2020
Published in: Applied Numerical Mathematics (Search for Journal in Brave)
Abstract: The choice of the parameter value for regularized inverse problems is critical to the results and remains a topic of interest. This article explores a criterion for selecting a good parameter value by maximizing the probability of the data, {{with no prior knowledge of the noise variance}}. These concepts are developed for and consequently regularization models by way of their Bayesian interpretations. Based on these concepts, an iterative scheme is proposed and demonstrated to converge accurately, and analytical convergence results are provided that substantiate these empirical observations. For some of the most common inverse problems, including MRI, SAR, denoising, and deconvolution, an extremely efficient algorithm is derived, making the iterative scheme very attractive for real case use. The computational concerns associated with the general case for any inverse problem are also carefully addressed. A robust set of 1D and 2D numerical simulations confirm the effectiveness of the proposed approach.
Full work available at URL: https://arxiv.org/abs/1812.11449
Recommendations
- Determination of regularization parameter via solving a multi-objective optimization problem
- Choosing regularization parameters in iterative methods for ill-posed problems
- A new choice rule for regularization parameters in Tikhonov regularization
- Using generalized cross validation to select regularization parameter for total variation regularization problems
- Regularization Parameter Selection in Total Variation Based Image Denoising
Image processing (compression, reconstruction, etc.) in information and communication theory (94A08) Computing methodologies for image processing (68U10) Numerical aspects of computer graphics, image analysis, and computational geometry (65D18)
Cites Work
- Nonlinear total variation based noise removal algorithms
- Some Comments on C P
- Estimation of the mean of a multivariate normal distribution
- Distributed optimization and statistical learning via the alternating direction method of multipliers
- Adapting to Unknown Smoothness via Wavelet Shrinkage
- Generalized Cross-Validation as a Method for Choosing a Good Ridge Parameter
- Pattern recognition and machine learning.
- The Little Engine that Could: Regularization by Denoising (RED)
- On the ``degrees of freedom of the lasso
- Robust uncertainty principles: exact signal reconstruction from highly incomplete frequency information
- Noise removal using fourth-order partial differential equation with applications to medical magnetic resonance images in space and time
- Augmented Lagrangian method, dual methods, and split Bregman iteration for ROF, vectorial TV, and high order models
- Generalized SURE for Exponential Families: Applications to Regularization
- Computational Methods for Inverse Problems
- Sparsity and incoherence in compressive sampling
- Regularization Parameter Selection for Nonlinear Iterative Image Restoration and MRI Reconstruction Using GCV and SURE-Based Methods
- Total generalized variation
- A Novel SURE-Based Criterion for Parametric PSF Estimation
- Tikhonov regularization and the L-curve for large discrete ill-posed problems
- Some large-scale matrix computation problems
- The projected GSURE for automatic parameter tuning in iterative shrinkage methods
- Regularization parameter estimation for large-scale Tikhonov regularization using a priori information
- Regularizing preconditioners by non-stationary iterated Tikhonov with general penalty term
- Iterated Tikhonov regularization with a general penalty term.
- Sampled Tikhonov regularization for large linear inverse problems
Cited In (7)
- Generalized Sparse Bayesian Learning and Application to Image Reconstruction
- Regularization of statistical inverse problems and the Bakushinskiĭ veto
- Sequential image recovery from noisy and under-sampled Fourier data
- Empirical Bayesian inference using a support informed prior
- Maximum likelihood estimation of regularization parameters in high-dimensional inverse problems: an empirical Bayesian approach. I: Methodology and experiments
- A semismooth Newton method for \(\mathrm{L}^1\) data fitting with automatic choice of regularization parameters and noise calibration
- Inverse problem solution and regularization parameter selection for current distribution reconstruction in switching arcs by inverting magnetic fields
Uses Software
This page was built for publication: Effective new methods for automated parameter selection in regularized inverse problems
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2301416)
