Geodesic
Iterative refinement of least squares solutions computed from normal equations
Mathematica Applicanda, Tome 17 (1989) no. 31, pp. 91-101.

Voir la notice de l'article provenant de la source Annales Societatis Mathematicae Polonae Series

It is shown that iterative refinement (using normal equations and exclusively standard floating point arithmetic with relative precision v) yields almost full accuracy of computed solution to regular linear least squares problem, provided some conditions.
DOI : 10.14708/ma.v17i31.1766
Classification : 65F10
Mots-clés : Iterative methods for linear systems 
@article{10_14708_ma_v17i31_1766,
     author = {Andrzej Kie{\l}basi\'nski},
     title = {Iterative refinement of least squares solutions computed from normal equations},
     journal = {Mathematica Applicanda},
     pages = { 91--101},
     publisher = {mathdoc},
     volume = {17},
     number = {31},
     year = {1989},
     doi = {10.14708/ma.v17i31.1766},
     language = {pl},
     url = {http://geodesic.mathdoc.fr/articles/10.14708/ma.v17i31.1766/}
}
TY  - JOUR
AU  - Andrzej Kiełbasiński
TI  - Iterative refinement of least squares solutions computed from normal equations
JO  - Mathematica Applicanda
PY  - 1989
SP  -  91
EP  - 101
VL  - 17
IS  - 31
PB  - mathdoc
UR  - http://geodesic.mathdoc.fr/articles/10.14708/ma.v17i31.1766/
DO  - 10.14708/ma.v17i31.1766
LA  - pl
ID  - 10_14708_ma_v17i31_1766
ER  - 
%0 Journal Article
%A Andrzej Kiełbasiński
%T Iterative refinement of least squares solutions computed from normal equations
%J Mathematica Applicanda
%D 1989
%P  91-101
%V 17
%N 31
%I mathdoc
%U http://geodesic.mathdoc.fr/articles/10.14708/ma.v17i31.1766/
%R 10.14708/ma.v17i31.1766
%G pl
%F 10_14708_ma_v17i31_1766
Andrzej Kiełbasiński. Iterative refinement of least squares solutions computed from normal equations. Mathematica Applicanda, Tome 17 (1989) no. 31, pp.  91-101. doi : 10.14708/ma.v17i31.1766. http://geodesic.mathdoc.fr/articles/10.14708/ma.v17i31.1766/

Cité par Sources :