Regularized Linear Inversion with Randomized Singular Value Decomposition
From MaRDI portal
Publication:5117954
DOI10.1007/978-3-030-48634-1_5zbMATH Open1451.65043arXiv1909.01947OpenAlexW3167796844MaRDI QIDQ5117954FDOQ5117954
Authors: Kazufumi Ito, Bangti Jin 
Publication date: 26 August 2020
Published in: Mathematical and Numerical Approaches for Multi-Wave Inverse Problems (Search for Journal in Brave)
Abstract: In this work, we develop efficient solvers for linear inverse problems based on randomized singular value decomposition (RSVD). This is achieved by combining RSVD with classical regularization methods, e.g., truncated singular value decomposition, Tikhonov regularization, and general Tikhonov regularization with a smoothness penalty. One distinct feature of the proposed approach is that it explicitly preserves the structure of the regularized solution in the sense that it always lies in the range of a certain adjoint operator. We provide error estimates between the approximation and the exact solution under canonical source condition, and interpret the approach in the lens of convex duality. Extensive numerical experiments are provided to illustrate the efficiency and accuracy of the approach.
Full work available at URL: https://arxiv.org/abs/1909.01947
Recommendations
- Regularization with randomized SVD for large-scale discrete inverse problems
- Randomized matrix approximation to enhance regularized projection schemes in inverse problems
- Optimal regularized low rank inverse approximation
- scientific article; zbMATH DE number 782089
- scientific article; zbMATH DE number 3974187
- Low complexity regularization of linear inverse problems
- Relaxed regularization for linear inverse problems
- Relaxed regularization for linear inverse problems
- A vector regularization method to solve linear inverse problems
- Radial basis function regularization for linear inverse problems with random noise
Numerical solutions to overdetermined systems, pseudoinverses (65F20) Ill-posedness and regularization problems in numerical linear algebra (65F22)
Cites Work
- Matrix Analysis
- Title not available (Why is that?)
- On the Perturbation of Pseudo-Inverses, Projections and Linear Least Squares Problems
- Finding structure with randomness: probabilistic algorithms for constructing approximate matrix decompositions
- Title not available (Why is that?)
- Fast monte-carlo algorithms for finding low-rank approximations
- Existence and Uniqueness for Electrode Models for Electric Current Computed Tomography
- Title not available (Why is that?)
- Title not available (Why is that?)
- A weighted pseudoinverse, generalized singular values, and constrained least squares problems
- The x-ray transform: singular value decomposition and resolution
- On the degree of ill-posedness of multi-dimensional magnetic particle imaging
- An analysis of finite element approximation in electrical impedance tomography
- Title not available (Why is that?)
- Tikhonov regularization and randomized GSVD
- An a posteriori parameter choice for Tikhonov regularization in the presence of modeling error
- On the decay rate of the singular values of bivariate functions
- Regularization of linear ill-posed problems with noisy right hand side and noisy operator
- Regularization with randomized SVD for large-scale discrete inverse problems
- A convergent adaptive finite element method for electrical impedance tomography
- Iterative Hessian sketch: fast and accurate solution approximation for constrained least-squares
- Random Projections for Classification: A Recovery Approach
- Randomized algorithms for large-scale inverse problems with general Tikhonov regularizations
- Accurate low-rank approximations via a few iterations of alternating least squares
- Subspace Iteration Randomization and Singular Value Problems
- Sketching meets random projection in the dual: a provable recovery algorithm for big and high-dimensional data
- Randomized algorithms for low-rank matrix factorizations: sharp performance bounds
- Inverse problems. Tikhonov theory and algorithms
- Modified truncated randomized singular value decomposition (MTRSVD) algorithms for large scale discrete ill-posed problems with general-form regularization
Cited In (10)
- Solution of the EEG inverse problem by random dipole sampling
- Solving, tracking and stopping streaming linear inverse problems
- Randomized matrix approximation to enhance regularized projection schemes in inverse problems
- On unifying randomized methods for inverse problems
- A probabilistic oracle inequality and quantification of uncertainty of a modified discrepancy principle for statistical inverse problems
- Randomized algorithms for large-scale inverse problems with general Tikhonov regularizations
- Regularization with randomized SVD for large-scale discrete inverse problems
- Compression approaches for the regularized solutions of linear systems from large-scale inverse problems
- Relaxed regularization for linear inverse problems
- A stochastic gradient descent approach with partitioned-truncated singular value decomposition for large-scale inverse problems of magnetic modulus data
This page was built for publication: Regularized Linear Inversion with Randomized Singular Value Decomposition
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5117954)
