Abstract
Estimates for the condition number of a matrix are useful in many areas of scientific computing, including: recursive least squares computations, optimization, eigenanalysis, and general nonlinear problems solved by linearization techniques where matrix modification techniques are used. The purpose of this paper is to propose anadaptiveLanczosestimator scheme, which we callale, for tracking the condition number of the modified matrix over time. Applications to recursive least squares (RLS) computations using the covariance method with sliding data windows are considered.ale is fast for relatively smalln-parameter problems arising in RLS methods in control and signal processing, and is adaptive over time, i.e., estimates at timet are used to produce estimates at timet+1. Comparisons are made with other adaptive and non-adaptive condition estimators for recursive least squares problems. Numerical experiments are reported indicating thatale yields a very accurate recursive condition estimator.
Similar content being viewed by others
References
S.T. Alexander,Adaptive Signal Processing (Springer Verlag, New York, 1986).
S.T. Alexander, C.-T. Pan and R.J. Plemmons, Analysis of a recursive least squares hyperbolic rotation scheme for signal processing, Lin. Alg. Appl. Special Issue on Electrical Eng. 98 (1988) 3–40.
K.J. Åström and B. Wittenmark,Adaptive Control (Addison-Wesley, Reading MA, 1989).
G.J. Bierman,Factorization Methods for Discrete Sequential Estimation (Academic Press, NY, 1977).
C.H. Bischof, Incremental condition estimation, SIAM J. on Matrix Anal. and Applic. 11 (1990) 644–659.
C.H. Bischof and G.M. Shroff, On updating signal subspaces, Tech. Rept., Argonne National Lab. (September 1989).
Å. Björck, Least Squares Methods, in:Handbook of Numerical Analysis, eds.. P. Ciarlet and J. Lions (Elsevier/North Holland, Amsterdam, 1989).
Å. Björck, Accurate downdating of least squares solutions, presented at the Annual SIAM Meeting, held at San Diego, CA (1989).
J.R. Bunch and C.P. Nielson, Updating the singular vaule decomposition, Numer. Math., 31 (1978) 111–129.
J.R. Bunch, C.P. Nielson and D.C. Sorensen, Rank-one modification of the symmetric eigenproblem, Numer. Math. 31 (1978) 31–48.
J.M. Cioffi, Limited precision effects in adaptive filtering, IEEE Trans. on CAS 34 (1987) 821–833.
J.M. Cioffi and T. Kailath, Windowed fast transversal filters, adaptive algorithms with normalization, IEEE Trans. on Acous., Speech and Sig. Proc. 33 (1985) 607–625.
P. Comon and G.H. Golub, Tracking a few extreme singular values and vectors in signal processing, Stanford Univ. Tech. Rept. NA-89-01 (1989).
R.D. DeGroat and R.A. Roberts, Highly parallel eigenvector update methods with applications to signal processing, SPIE Adv. Alg. and Arc. for Signal Proc. 696 (1986) 62–70.
P.E. Gill, G.H. Golub, W. Murray and M.A. Saunders, Methods for modifying matrix factorizations, Math. Comput. 28 (1974) 505–535.
G.H. Golub, Matrix decompositions and statistical calculations, in:Satistical Computation eds. R.C. Milton and J.A. Nelder (Academic Press, New York, 1969) 365–395.
G.H. Golub, Some modified matrix eigenvalue problems, SIAM Review 15 (1973) 318–334.
G.H. Golub and W. Kahan, Calculating the singular values and pseudo inverse of a matrix, SIAM J. Num. Anal. Ser. B 2 (1965) 205–224.
G.H. Golub, F.T. Luk and M.L. Overton, A block Lanczos method for computing the singular values and corresponding singular vectors of a matrix, ACM Transcation on Mathematical Software 7 (1981) 149–169.
G.H. Golub and C. Van Loan,Matrix Computations (Johns Hopkins Press, Baltimore, 1989).
S. Haykin,Adaptive Filter Theory (Prentice-Hall, Englewood Cliffs, NJ, 1986).
C.S. Henkel and R.J. Plemmons, Recursive least squares computations on a hypercube multiprocessor using the covariance factorization, SIAM J. Sci. Statist. Comp. 12 (1991) 95–106.
N.J. Higham, A survey of condition estimators for triangular matrices, SIAM Review 29 (1987) 575–596.
Y.H. Hu, Adaptive methods for real time Pisarenko spectrum estimation,Proc. ICASSP, Tampa, FL (1985).
C. Lanczos, An iteration method for the solution of the eigenvalue problem of linear differential and integral operators. J. Res. Nat. Bur. Stand. 45 (1950) 255–281.
F. Milinazzo, C. Zala and I. Barrodale, On the rate of growth of condition numbers for convolution matrices, IEEE Trans. Acous., Speech and Sig. Proc. 35 (1987) 471–475.
C. Moler, J. Little and S. Bangert,Pro-Matlab Users Guide (The Mathworks, Sherborn, MA, 1987).
M. Moonen, P. van Dooren and J. Vandewalle, Parallel updating algorithms, part one: The ordinary singular value decomposition, preprint (1989).
M. Morf and T. Kailath, Square root algorithms and least squares estimation, IEEE Trans. on Automatic Control 20 (1975) 487–497.
C.-T. Pan and R.J. Plemmons, Least squares modifications with inverse factorizations: parallel implications, Computational and Applied Math. 27 (1989) 109–127.
D.J. Pierce and R.J. Plemmons, Fast adaptive condition estimation, SIAM J. on Matrix Anal. and Applic., to appear 1991.
D.J. Pierce and R.J. Plemmons, Tracking the condition number for RLS in signal processing, Mathematics of Control, Signals and Systems, to appear 1991.
R.J. Plemmons, Recursive least squares computations,Proc. Inter. Conf. on Math. in Networks and Syst. (Birkhauser Boston, Inc.) 3 (1990) 495–502.
S. Sastry and M. Bodson,Adaptive Control: Stability, Convergence, and Robustness (Prentice-Hall, Englewood Cliffs, NJ, 1989).
R.C. Thompson, The behavior of eigenvalues and singular values under perturbations of restricted rank, Linear Algebra and Its Applications 13 (1976) 69–78.
Author information
Authors and Affiliations
Additional information
Research supported by the US Air Force under grant no. AFOSR-88-0285.
Research supported by the US Army under grant no. DAAL03-90-G-105.
Research supported by the US Air Force under grant no. AFOSR-88-0285.
Rights and permissions
About this article
Cite this article
Ferng, W.R., Golub, G.H. & Plemmons, R.J. Adaptive Lanczos methods for recursive condition estimation. Numer Algor 1, 1–19 (1991). https://doi.org/10.1007/BF02145580
Received:
Issue Date:
DOI: https://doi.org/10.1007/BF02145580