LU Preconditioning for Overdetermined Sparse Least Squares Problems

Abstract : We investigate how to use an LU factorization with the classical LSQR routine for solving overdetermined sparse least squares problems. Usually L is much better conditioned than A and iterating with L instead of A results in faster convergence. When a runtime test indicates that L is not sufficiently well-conditioned, a partial orthogonalization of L accelerates the convergence. Numerical experiments illustrate the good behavior of our algorithm in terms of storage and convergence.
Type de document :
Communication dans un congrès
11th International Conference on Parallel Processing and Applied Mathematics (PPAM 2015), Sep 2015, Krakow, Poland. 2015, Lecture Notes in Computer Science. 〈http://ppam.pl〉
Liste complète des métadonnées

https://hal.inria.fr/hal-01223069
Contributeur : Marc Baboulin <>
Soumis le : dimanche 1 novembre 2015 - 20:51:18
Dernière modification le : jeudi 11 janvier 2018 - 06:25:42

Identifiants

  • HAL Id : hal-01223069, version 1

Citation

Gary Howell, Marc Baboulin. LU Preconditioning for Overdetermined Sparse Least Squares Problems. 11th International Conference on Parallel Processing and Applied Mathematics (PPAM 2015), Sep 2015, Krakow, Poland. 2015, Lecture Notes in Computer Science. 〈http://ppam.pl〉. 〈hal-01223069〉

Partager

Métriques

Consultations de la notice

98