Thesis (Selection of subject)Thesis (Selection of subject)(version: 368)
Thesis details
   Login via CAS
Inexact preconditioning for iterative methods
Thesis title in Czech: Inexact preconditioning for iterative methods
Thesis title in English: Inexact preconditioning for iterative methods
Key words: iterační metody pro řešení soustav lineárních algebraických rovnic, výkonné počítačové architektury, zpětná stabilita
English key words: iterative methods, linear algebraic equations, High-Performance computing, preconditioning, numerical linear algebra, finite precision
Academic year of topic announcement: 2019/2020
Thesis type: dissertation
Thesis language:
Department: Department of Numerical Mathematics (32-KNM)
Supervisor: Erin Claire Carson, Ph.D.
Author: hidden - assigned and confirmed by the Study Dept.
Date of registration: 20.02.2020
Date of assignment: 20.02.2020
Confirmed by Study dept. on: 13.10.2020
Guidelines
We consider the use of Krylov subspace methods for solving square linear systems Ax = b
in an iterative fashion. It is standard practice to use preconditioning in order to reduce the
number of iterations, which involves solving a transformed linear system with theoretically
better numerical properties.

There is a complicated tradeoff between the cost per iteration associated with using a particular preconditioner and the resulting convergence rate. In general, we expect the cost per iteration to
increase as the number of iterations decrease. Investigating this complicated tradeoff for various iterative methods and applications, in both theory and practice, will be the goal of this project.
References
Y. Saad. ILUT: A dual threshold incomplete LU factorization. Numer. Lin. Alg. Appl.,
1(4):387{402, 1994.

J. Liesen and Z. Strakos. Krylov subspace methods: principles and analysis. Oxford University
Press, 2012.

A. Greenbaum. Behavior of slightly perturbed Lanczos and conjugate-gradient recurrences.
Lin. Alg. Appl., 113:7-63, 1989.

Rozložník, Miroslav. Saddle-point problems and their iterative solution. Birkhäuser, 2018.

Benzi, Michele. "Preconditioning techniques for large linear systems: a survey." Journal of computational Physics 182.2 (2002): 418-477.

Saad, Yousef. Iterative methods for sparse linear systems. Vol. 82. SIAM, 2003.

Elman, Howard C. "A stability analysis of incomplete LU factorizations." Mathematics of Computation (1986): 191-217.
Preliminary scope of work in English
The goal of the work is to study the convergence and stability properties of iterative methods subject to inexactness in preconditioning and design of inexact preconditioning techniques suitable for high-performance computing.
 
Charles University | Information system of Charles University | http://www.cuni.cz/UKEN-329.html