Regularization of large scale total least squares problems

Publikationstyp
Conference Paper
Date Issued
2011
Sprache
English
Author(s)
Voss, Heinrich 
Lampe, Jörg  
Institut
Numerische Simulation E-10 (H)  
TORE-URI
http://hdl.handle.net/11420/11773
First published in
Communications in computer and information science  
Number in series
245
Start Page
22
End Page
37
Citation
Communications in Computer and Information Science 245: 22-37 (2011)
Contribution to Conference
10th International Conference on Computer Information Systems and Industrial Management Applications, CISIM 2011  
Publisher DOI
10.1007/978-3-642-27245-5_6
Scopus ID
2-s2.0-83755171556
Publisher
Springer
ISBN of container
978-3-64-227244-8
The total least squares (TLS) method is an appropriate approach for linear systems when not only the right-hand side but also the system matrix is contaminated by some noise. For ill-posed problems regularization is necessary to stabilize the computed solutions. In this presentation we discuss two approaches for regularizing large scale TLS problems. One which is based on adding a quadratic constraint and a Tikhonov type regularization concept.
DDC Class
510: Mathematik