Mixed and componentwise condition numbers for a linear function of the solution of the total least squares problem

被引:16
作者
Diao, Huai-An [1 ]
Sun, Yang [1 ]
机构
[1] Northeast Normal Univ, Sch Math & Stat, 5268 Renmin St, Changchun 130024, Jilin, Peoples R China
关键词
Total least squares problem; Componentwise perturbation; Condition number; Adjoint operator; Structured perturbation; RESTARTED LANCZOS BIDIAGONALIZATION; STRUCTURED CONDITION NUMBERS; SMALLEST SINGULAR TRIPLETS; PERTURBATION ANALYSIS; MATRICES; SYSTEMS; SENSITIVITY; EQUATIONS;
D O I
10.1016/j.laa.2018.01.008
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
In this paper, we consider the mixed and componentwise condition numbers for a linear function Lx of the solution to the total least squares (TLS) problem. We derive the explicit expressions of the mixed and componentwise condition numbers through the dual techniques under both unstructured and structured componentwise perturbations. The sharp upper bounds for condition numbers are obtained. An efficient condition estimation algorithm is proposed, which can be integrated into the iterative method for solving large scale TLS problems. Moreover, the new derived condition number expressions can recover the previous results on the condition analysis for the TLS problem when L = I-n. Numerical experiments show the effectiveness of the introduced condition numbers and condition estimation algorithm. (C) 2018 Elsevier Inc. All rights reserved.
引用
收藏
页码:1 / 29
页数:29
相关论文
共 40 条
[1]  
[Anonymous], 2002, Accuracy and stability of numerical algorithms
[2]   A partial condition number for linear least squares problems [J].
Arioli, Mario ;
Baboulin, Marc ;
Gratton, Serge .
SIAM JOURNAL ON MATRIX ANALYSIS AND APPLICATIONS, 2007, 29 (02) :413-433
[3]   A CONTRIBUTION TO THE CONDITIONING OF THE TOTAL LEAST-SQUARES PROBLEM [J].
Baboulin, Marc ;
Gratton, Serge .
SIAM JOURNAL ON MATRIX ANALYSIS AND APPLICATIONS, 2011, 32 (03) :685-699
[4]   Using dual techniques to derive componentwise and mixed condition numbers for a linear function of a linear least squares solution [J].
Baboulin, Marc ;
Gratton, Serge .
BIT NUMERICAL MATHEMATICS, 2009, 49 (01) :3-19
[5]   Augmented implicitly restarted Lanczos bidiagonalization methods [J].
Baglama, J ;
Reichel, L .
SIAM JOURNAL ON SCIENTIFIC COMPUTING, 2005, 27 (01) :19-42
[6]  
Björck Å, 2000, SIAM J MATRIX ANAL A, V22, P413, DOI 10.1137/S0895479899355414
[7]   ON THE SENSITIVITY OF SOLUTION COMPONENTS IN LINEAR-SYSTEMS OF EQUATIONS [J].
CHANDRASEKARAN, S ;
IPSEN, ICF .
SIAM JOURNAL ON MATRIX ANALYSIS AND APPLICATIONS, 1995, 16 (01) :93-112
[8]   Mixed and componentwise condition numbers for rectangular structured matrices [J].
Cucker, Felipe ;
Diao, Huaian .
CALCOLO, 2007, 44 (02) :89-115
[9]   On mixed and componentwise condition numbers for Moore-Penrose inverse and linear least squares problems [J].
Cucker, Felipe ;
Diao, Huaian ;
Wei, Yimin .
MATHEMATICS OF COMPUTATION, 2007, 76 (258) :947-963
[10]  
Diao H.-A., 2016, ARXIV161203645