Smoothing and regularization with modified sparse approximate inverses (Q543017): Difference between revisions
From MaRDI portal
Created a new Item |
ReferenceBot (talk | contribs) Changed an Item |
||
(10 intermediate revisions by 7 users not shown) | |||
Property / author | |||
Property / author: Q543015 / rank | |||
Property / author | |||
Property / author: Matous Sedlacek / rank | |||
Property / review text | |||
Summary: Sparse approximate inverses \(M\) which satisfy \(\min_{M}\| AM-I\|_{F}\) have shown to be an attractive alternative to classical smoothers like Jacobi or Gauss-Seidel [\textit{W.-P. Tang} and \textit{W. L. Wan}, SIAM J. Matrix Anal. Appl. 21, No. 4, 1236-1252 (2000; Zbl 1049.65146)]. The static and dynamic computation of a SAI and a SPAI [\textit{M. J. Grote} and \textit{T. Huckle}, SIAM J. Sci. Comput. 18, No. 3, 838--853 (1997; Zbl 0872.65031)], respectively, comes along with advantages like inherent parallelism and robustness with equal smoothing properties. Here, we are interested in developing preconditioners that can incorporate probing conditions for improving the approximation relative to high- or low-frequency subspaces. We present analytically derived optimal smoothers for the discretization of the constant-coefficient Laplace operator. On this basis, we introduce probing conditions in the generalized Modified SPAI (MSPAI) approach [\textit{T. Huckle} and \textit{A. Kallischko}, Int. J. Comput. Math. 84, No. 8, 1225--1248 (2007; Zbl 1126.65035)] which yields efficient smoothers for multigrid. In the second part, we transfer our approach to the domain of ill-posed problems to recover original information from blurred signals. Using the probing facility of MSPAI, we impose the preconditioner to act as approximately zero on the noise subspace. In combination with an iterative regularization method, it thus becomes possible to reconstruct the original information more accurately in many cases. A variety of numerical results demonstrate the usefulness of this approach. | |||
Property / review text: Summary: Sparse approximate inverses \(M\) which satisfy \(\min_{M}\| AM-I\|_{F}\) have shown to be an attractive alternative to classical smoothers like Jacobi or Gauss-Seidel [\textit{W.-P. Tang} and \textit{W. L. Wan}, SIAM J. Matrix Anal. Appl. 21, No. 4, 1236-1252 (2000; Zbl 1049.65146)]. The static and dynamic computation of a SAI and a SPAI [\textit{M. J. Grote} and \textit{T. Huckle}, SIAM J. Sci. Comput. 18, No. 3, 838--853 (1997; Zbl 0872.65031)], respectively, comes along with advantages like inherent parallelism and robustness with equal smoothing properties. Here, we are interested in developing preconditioners that can incorporate probing conditions for improving the approximation relative to high- or low-frequency subspaces. We present analytically derived optimal smoothers for the discretization of the constant-coefficient Laplace operator. On this basis, we introduce probing conditions in the generalized Modified SPAI (MSPAI) approach [\textit{T. Huckle} and \textit{A. Kallischko}, Int. J. Comput. Math. 84, No. 8, 1225--1248 (2007; Zbl 1126.65035)] which yields efficient smoothers for multigrid. In the second part, we transfer our approach to the domain of ill-posed problems to recover original information from blurred signals. Using the probing facility of MSPAI, we impose the preconditioner to act as approximately zero on the noise subspace. In combination with an iterative regularization method, it thus becomes possible to reconstruct the original information more accurately in many cases. A variety of numerical results demonstrate the usefulness of this approach. / rank | |||
Normal rank | |||
Property / Mathematics Subject Classification ID | |||
Property / Mathematics Subject Classification ID: 65F08 / rank | |||
Normal rank | |||
Property / Mathematics Subject Classification ID | |||
Property / Mathematics Subject Classification ID: 65F10 / rank | |||
Normal rank | |||
Property / Mathematics Subject Classification ID | |||
Property / Mathematics Subject Classification ID: 65F35 / rank | |||
Normal rank | |||
Property / Mathematics Subject Classification ID | |||
Property / Mathematics Subject Classification ID: 65Y05 / rank | |||
Normal rank | |||
Property / zbMATH DE Number | |||
Property / zbMATH DE Number: 5910099 / rank | |||
Normal rank | |||
Property / author | |||
Property / author: Thomas Kilian Huckle / rank | |||
Normal rank | |||
Property / author | |||
Property / author: Matous Sedlacek / rank | |||
Normal rank | |||
Property / describes a project that uses | |||
Property / describes a project that uses: RestoreTools / rank | |||
Normal rank | |||
Property / describes a project that uses | |||
Property / describes a project that uses: MSPAI / rank | |||
Normal rank | |||
Property / describes a project that uses | |||
Property / describes a project that uses: Regularization tools / rank | |||
Normal rank | |||
Property / MaRDI profile type | |||
Property / MaRDI profile type: MaRDI publication profile / rank | |||
Normal rank | |||
Property / full work available at URL | |||
Property / full work available at URL: https://doi.org/10.1155/2010/930218 / rank | |||
Normal rank | |||
Property / OpenAlex ID | |||
Property / OpenAlex ID: W2054744919 / rank | |||
Normal rank | |||
Property / Wikidata QID | |||
Property / Wikidata QID: Q58652014 / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Frobenius norm minimization and probing for preconditioning / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Q3968379 / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Approximate inverse preconditionings for sparse linear systems / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Parallel Preconditioning with Sparse Approximate Inverses / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Sparse Approximate Inverses and Target Matrices / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Iterative Solution Methods / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Q3825995 / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: The Interface Probing Technique in Domain Decomposition / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: A comparative study of sparse approximate inverse preconditioners / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Q3752135 / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Compact Fourier Analysis for Designing Multigrid Methods / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Sparse Approximate Inverse Smoother for Multigrid / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Robust Parallel Smoothing for Multigrid Via Sparse Approximate Inverses / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Q5515676 / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Q4895893 / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Q4524533 / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Q4296893 / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Restoring Images Degraded by Spatially Variant Blur / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Regularization tools: A Matlab package for analysis and solution of discrete ill-posed problems / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Iterative methods for image deblurring: A Matlab object-oriented approach / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Analysis of Discrete Ill-Posed Problems by Means of the L-Curve / rank | |||
Normal rank | |||
links / mardi / name | links / mardi / name | ||
Latest revision as of 05:19, 4 July 2024
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Smoothing and regularization with modified sparse approximate inverses |
scientific article |
Statements
Smoothing and regularization with modified sparse approximate inverses (English)
0 references
20 June 2011
0 references
Summary: Sparse approximate inverses \(M\) which satisfy \(\min_{M}\| AM-I\|_{F}\) have shown to be an attractive alternative to classical smoothers like Jacobi or Gauss-Seidel [\textit{W.-P. Tang} and \textit{W. L. Wan}, SIAM J. Matrix Anal. Appl. 21, No. 4, 1236-1252 (2000; Zbl 1049.65146)]. The static and dynamic computation of a SAI and a SPAI [\textit{M. J. Grote} and \textit{T. Huckle}, SIAM J. Sci. Comput. 18, No. 3, 838--853 (1997; Zbl 0872.65031)], respectively, comes along with advantages like inherent parallelism and robustness with equal smoothing properties. Here, we are interested in developing preconditioners that can incorporate probing conditions for improving the approximation relative to high- or low-frequency subspaces. We present analytically derived optimal smoothers for the discretization of the constant-coefficient Laplace operator. On this basis, we introduce probing conditions in the generalized Modified SPAI (MSPAI) approach [\textit{T. Huckle} and \textit{A. Kallischko}, Int. J. Comput. Math. 84, No. 8, 1225--1248 (2007; Zbl 1126.65035)] which yields efficient smoothers for multigrid. In the second part, we transfer our approach to the domain of ill-posed problems to recover original information from blurred signals. Using the probing facility of MSPAI, we impose the preconditioner to act as approximately zero on the noise subspace. In combination with an iterative regularization method, it thus becomes possible to reconstruct the original information more accurately in many cases. A variety of numerical results demonstrate the usefulness of this approach.
0 references
0 references