EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

Krylov Subspace Based FISTA‐Type Methods for Linear Discrete Ill‐Posed Problems

Alessandro Buccini, Fei Chen, Mirjeta Pasha, Lothar Reichel

Numerical Linear Algebra with Applications · 2025-02 · 32 (1)

DOI: 10.1002/nla.2610 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Numerical Linear Algebra with Applications, online erschienen am 29.12.2024. Verfasst von Alessandro Buccini und 3 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 36 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: Gruppo Nazionale per il Calcolo Scientifico, Fondazione di Sardegna, National Science Foundation.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Several iterative soft‐thresholding algorithms, such as FISTA, have been proposed in the literature for solving regularized linear discrete inverse problems that arise in various applications in science and engineering. These algorithms are easy to implement, but their rates of convergence may be slow. This paper describes novel approaches to reduce the computations required for each iteration by using Krylov subspace techniques. Specifically, we propose to impose sparsity on the coefficients in the representation of the computed solution in terms of a Krylov subspace basis. Several numerical examples from image deblurring and computerized tomography are used to illustrate the efficiency and accuracy of the proposed methods.

Förderung

  • Gruppo Nazionale per il Calcolo Scientifico
  • Fondazione di Sardegna
  • National Science Foundation
  • Ministero dell'Università e della Ricerca

Zitieren

APA

Buccini, A., Chen, F., Pasha, M., & Reichel, L. (2025). Krylov Subspace Based FISTA‐Type Methods for Linear Discrete Ill‐Posed Problems. Numerical Linear Algebra with Applications, 32(1). https://doi.org/10.1002/nla.2610

Vancouver

Buccini A, Chen F, Pasha M, Reichel L. Krylov Subspace Based FISTA‐Type Methods for Linear Discrete Ill‐Posed Problems. Numerical Linear Algebra with Applications. 2025;32(1). doi:10.1002/nla.2610

BibTeX
@article{Buccini2025,
  author = {Buccini, Alessandro and Chen, Fei and Pasha, Mirjeta and Reichel, Lothar},
  title = {Krylov Subspace Based FISTA‐Type Methods for Linear Discrete Ill‐Posed Problems},
  journal = {Numerical Linear Algebra with Applications},
  year = {2025},
  volume = {32},
  number = {1},
  doi = {10.1002/nla.2610}
}
Kostenlos starten →