A class of optimized row projection methods for solving large non-symmetric linear systems : Notas de Matemática, 74

We present in this paper optimal and accelerated row projection algorithms arising from new results that allow us to define the iterate xfc+1 as the projection of xk onto a hyperplane which minimizes its distance to the solution x*. These algorithms also use a novel partition strategy into blocks ba...

Descripción completa

Detalles Bibliográficos
Autores principales: Scolnik, Hugo Daniel, Echebest, Nélida Ester, Guardarucci, María Teresa, Vacchino, María Cristina
Formato: Publicacion seriada
Lenguaje:Inglés
Publicado: 2000
Materias:
Matemática
Projected aggregate methods
row partition strategies
parallel iterative methods
Acceso en línea:http://sedici.unlp.edu.ar/handle/10915/172779
Aporte de:
SEDICI (UNLP) de Universidad Nacional de La Plata
id I19-R120-10915-172779
record_format dspace
spelling I19-R120-10915-1727792024-11-08T04:10:41Z http://sedici.unlp.edu.ar/handle/10915/172779 A class of optimized row projection methods for solving large non-symmetric linear systems : Notas de Matemática, 74 Scolnik, Hugo Daniel Echebest, Nélida Ester Guardarucci, María Teresa Vacchino, María Cristina 2000 2024-11-07T16:26:26Z en Matemática Projected aggregate methods row partition strategies parallel iterative methods We present in this paper optimal and accelerated row projection algorithms arising from new results that allow us to define the iterate xfc+1 as the projection of xk onto a hyperplane which minimizes its distance to the solution x*. These algorithms also use a novel partition strategy into blocks based on sequential estimations of their condition numbers. Numerical results are given showing the new algorithms are more robust than Krylov subspace based methods, although the latter are generally faster when they converge. Material digitalizado en SEDICI gracias a la colaboración de la Biblioteca del Departamento de Matemática de la Facultad de Ciencias Exactas (UNLP). Facultad de Ciencias Exactas Publicacion seriada Publicacion seriada http://creativecommons.org/licenses/by-nc-sa/4.0/ Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0) application/pdf
institution Universidad Nacional de La Plata
institution_str I-19
repository_str R-120
collection SEDICI (UNLP)
language Inglés
topic Matemática
Projected aggregate methods
row partition strategies
parallel iterative methods
spellingShingle Matemática
Projected aggregate methods
row partition strategies
parallel iterative methods
Scolnik, Hugo Daniel
Echebest, Nélida Ester
Guardarucci, María Teresa
Vacchino, María Cristina
A class of optimized row projection methods for solving large non-symmetric linear systems : Notas de Matemática, 74
topic_facet Matemática
Projected aggregate methods
row partition strategies
parallel iterative methods
description We present in this paper optimal and accelerated row projection algorithms arising from new results that allow us to define the iterate xfc+1 as the projection of xk onto a hyperplane which minimizes its distance to the solution x*. These algorithms also use a novel partition strategy into blocks based on sequential estimations of their condition numbers. Numerical results are given showing the new algorithms are more robust than Krylov subspace based methods, although the latter are generally faster when they converge.
format Publicacion seriada
Publicacion seriada
author Scolnik, Hugo Daniel
Echebest, Nélida Ester
Guardarucci, María Teresa
Vacchino, María Cristina
author_facet Scolnik, Hugo Daniel
Echebest, Nélida Ester
Guardarucci, María Teresa
Vacchino, María Cristina
author_sort Scolnik, Hugo Daniel
title A class of optimized row projection methods for solving large non-symmetric linear systems : Notas de Matemática, 74
title_short A class of optimized row projection methods for solving large non-symmetric linear systems : Notas de Matemática, 74
title_full A class of optimized row projection methods for solving large non-symmetric linear systems : Notas de Matemática, 74
title_fullStr A class of optimized row projection methods for solving large non-symmetric linear systems : Notas de Matemática, 74
title_full_unstemmed A class of optimized row projection methods for solving large non-symmetric linear systems : Notas de Matemática, 74
title_sort class of optimized row projection methods for solving large non-symmetric linear systems : notas de matemática, 74
publishDate 2000
url http://sedici.unlp.edu.ar/handle/10915/172779
work_keys_str_mv AT scolnikhugodaniel aclassofoptimizedrowprojectionmethodsforsolvinglargenonsymmetriclinearsystemsnotasdematematica74
AT echebestnelidaester aclassofoptimizedrowprojectionmethodsforsolvinglargenonsymmetriclinearsystemsnotasdematematica74
AT guardaruccimariateresa aclassofoptimizedrowprojectionmethodsforsolvinglargenonsymmetriclinearsystemsnotasdematematica74
AT vacchinomariacristina aclassofoptimizedrowprojectionmethodsforsolvinglargenonsymmetriclinearsystemsnotasdematematica74
AT scolnikhugodaniel classofoptimizedrowprojectionmethodsforsolvinglargenonsymmetriclinearsystemsnotasdematematica74
AT echebestnelidaester classofoptimizedrowprojectionmethodsforsolvinglargenonsymmetriclinearsystemsnotasdematematica74
AT guardaruccimariateresa classofoptimizedrowprojectionmethodsforsolvinglargenonsymmetriclinearsystemsnotasdematematica74
AT vacchinomariacristina classofoptimizedrowprojectionmethodsforsolvinglargenonsymmetriclinearsystemsnotasdematematica74
_version_ 1827178669796229120

Ejemplares similares