×

zbMATH — the first resource for mathematics

The element-wise weighted total least-squares problem. (English) Zbl 1429.62298
Summary: A new technique is considered for parameter estimation in a linear measurement error model \(AX\approx B, A = A_{0}+ \tilde A B = B_0+ \tilde B, A_0 X_{0}=B_{0}\) with row-wise independent and non-identically distributed measurement errors \(\tilde A, \tilde B\). Here, \(A_{0}\) and \(B_{0}\) are the true values of the measurements \(A\) and \(B\), and \(X_{0}\) is the true value of the parameter \(X\). The total least-squares method yields an inconsistent estimate of the parameter in this case. Modified total least-squares problem, called element-wise weighted total least-squares, is formulated so that it provides a consistent estimator, i.e., the estimate \(\hat X\) converges to the true value \(X_{0}\) as the number of measurements increases. The new estimator is a solution of an optimization problem with the parameter estimate \(\hat X\) and the correction \(\Delta D = [\Delta A\Delta B]\), applied to the measured data \(D=[AB]\), as decision variables. An equivalent unconstrained problem is derived by minimizing analytically over the correction \(\Delta D\), and an iterative algorithm for its solution, based on the first order optimality condition, is proposed. The algorithm is locally convergent with linear convergence rate. For large sample size the convergence rate tends to quadratic.

MSC:
62J05 Linear regression; mixed models
62-08 Computational methods for problems pertaining to statistics
65F20 Numerical solutions to overdetermined systems, pseudoinverses
PDF BibTeX XML Cite
Full Text: DOI
References:
[1] Cheng, C.; Van Ness, J.W., Statistical regression with measurement error, (1999), Arnold London · Zbl 0947.62046
[2] De Moor, B., Structured total least squares and \(L_2\) approximation problems, Linear algebra appl., 188-189, 163-207, (1993) · Zbl 0781.65028
[3] Fisher, G.W., Matrix analysis of metamorphic mineral assemblages and reactions, Contrib. mineral. petrol., 102, 69-77, (1989)
[4] Fuller, W.A., Measurement error models, (1987), Wiley New York · Zbl 0800.62413
[5] Gallo, P.P., Consistency of regression estimates when some variables are subject to error, Comm. statist. B-theory methods, 11, 973-983, (1982) · Zbl 0515.62064
[6] Golub, G.H.; Van Loan, C.F., An analysis of the total least squares problem, SIAM J. numer. anal., 17, 883-893, (1980) · Zbl 0468.65011
[7] Kukush, A.; Van Huffel, S., Consistency of elementwise-weighted total least squares estimator in a multivariate errors-in-variables model \(\mathit{AX} = B\), Metrika, 59, 1, 75-97, (2004) · Zbl 1062.62100
[8] Kukush, A., Markovsky, I., Van Huffel, S., 2002. About the convergence of the computational algorithm for the EW-TLS estimator. Technical Report 02-49, Department of EE, K.U. Leuven. Available at:
[9] Premoli, A.; Rastello, M.L., The parametric quadratic form method for solving TLS problems with elementwise weighting, (), 67-76 · Zbl 0995.65011
[10] Sprent, P., A generalized least-squares approach to linear functional relationships, J. roy. statist. soc. B, 28, 278-297, (1966) · Zbl 0147.37804
[11] Van Huffel, S. (Ed.), 1997. Recent Advances in Total Least Squares Techniques and Errors-in-Variables Modeling. SIAM, Philadelphia. · Zbl 0861.00018
[12] Van Huffel, S., Lemmerling, P. (Eds.), 2002. Total Least Squares and Errors-in-Variables Modeling: Analysis, Algorithms and Applications. Kluwer Academic Publishers, Dordrecht, The Netherlands. · Zbl 0984.00011
[13] Van Huffel, S.; Vandewalle, J., Analysis and properties of the generalized total least squares problem \(\mathit{AX} \approx B\) when some or all columns in \(A\) are subject to error, SIAM J. matrix anal., 10, 3, 294-315, (1989) · Zbl 0681.65025
[14] Van Huffel, S.; Vandewalle, J., The total least squares problemcomputational aspects and analysis, (1991), SIAM Philadelphia
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.