Abstract
The truncated singular value decomposition (SVD) is considered as a method for regularization of ill-posed linear least squares problems. In particular, the truncated SVD solution is compared with the usual regularized solution. Necessary conditions are defined in which the two methods will yield similar results. This investigation suggests the truncated SVD as a favorable alternative to standard-form regularization in cases of ill-conditioned matrices with well-determined numerical rank.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
H. C. Andrews and B. R. Hunt,Digital Image Restoration, Prentice-Hall (1977).
A. Ben-Israel and T. N. E. Greville,Generalized Inverses: Theory and Applications, Wiley-Interscience (1974).
T. F. Chan and D. Foulser,Effective condition numbers for linear systems, Tech Memo 86-05, Saxpy Computer Corporation, Sunnyvale (1986).
T. F. Chan and P. C. Hansen,Computing truncated SVD least squares solutions by rank revealing QR-factorizations, CAM Report 86-01, Dept. of Mathematics, U.C.L.A. (1986).
U. Eckhardt and K. Mika,Numerical treatment of incorrectly posed problems — a case study; in J. Albrecht and L. Collatz (Eds.),Numerical Treatment of Integral Equations, Workshop on numerical treatment of integral equations, Oberwolfach. November 18–24, 1979, Birkhäuser Verlag (1980), pp. 92–101.
L. Eldén,Algorithms for regularization of ill-conditioned least squares problems, BIT 17 (1977), 134–145.
L. Eldén,The numerical solution of a non-characteristic Cauchy problem for a parabolic equation; in P. Deuflhard and E. Hairer (Eds.),Numerical Treatment of Inverse Problems in Differential and Integral Equations, Birkhäuser Verlag (1983), pp. 246–268.
G. E. Forsythe, M. A. Malcolm and C. B. Moler,Computer Methods for Mathematical Computations, Prentice-Hall (1977).
G. H. Golub, V. Klema and G. W. Stewart,Rank degeneracy and least squares problems, Technical Report TR-456, Computer Science Department, University of Maryland (1976).
G. H. Golub and C. F. Van Loan,Matrix Computations, North Oxford Academic (1983).
P. C. Hansen and S. Christiansen,An SVD analysis of linear algebraic equations derived from first kind integral equations, J. Comp. Appl. Math. 12&13 (1985), 341–357.
R. J. Hanson,A numerical method for solving Fredholm integral equations of the first kind using singular values, SIAM J. Numer. Anal. 8 (1971), 616–622.
C. L. Lawson and R. J. Hanson,Solving Least Squares Problems, Prentice Hall (1974).
A. K. Louis and F. Natterer,Mathematical problems of computerized tomography, Proc. IEEE 71 (1983), 379–389.
B. C. Moore and A. J. Laub,Computation of supremal (A,B)-invariant and controllability subspaces, IEEE Trans. Automat. Contr. AC-23 (1978), 783–792.
F. Natterer,Numerical inversion of the Radon transform, Numer. Math. 30 (1978), 81–91.
D. L. Phillips,A technique for the numerical solution of certain integral equations of the first kind, J. ACM 9 (1962), 84–97.
A. N. Tikhonov,Solution of incorrectly formulated problems and the regularization method, Dokl. Akad. Nauk. SSSR 151 (1963), 501–504 = Soviet Math. Dokl. 4 (1963), 1035–1038.
D. W. Tufts and R. Kumaresan,Singular value decomposition and improved frequency estimation using linear prediction, IEEE Trans. Acoust., Speech, Signal Processing ASSP-30 (1982), 671–675.
P. M. Van Dooren,The generalized eigenstructure problem in linear system theory, IEEE Trans. Automat. Contr. AC-26 (1981), 111–129.
J. M. Varah,On the numerical solution of ill-conditioned linear systems with applications to ill-posed problems, SIAM J. Numer. Anal. 10 (1973), 257–267.
J. M. Varah,A practical examination of some numerical methods for linear discrete ill-posed problems, SIAM Review 21 (1979), 100–111.
P.-Å. Wedin,Perturbation bounds in connection with the singular value decomposition, BIT 12 (1972), 99–111.
P.-Å. Wedin,Perturbation theory for pseudo-inverses, BIT 13 (1973), 217–232.
P.-Å. Wedin,On the almost rank deficient case of the least squares problem, BIT 13 (1973), 344–354.
Author information
Authors and Affiliations
Additional information
This work was carried out while the author visited the Dept. of Computer Science, Stanford University, California, U.S.A., and was supported in part by National Science Foundation Grant Number DCR 8412314, by a Fulbright Supplementary Grant, and by the Danish Space Board.
Rights and permissions
About this article
Cite this article
Hansen, P.C. The truncatedSVD as a method for regularization. BIT 27, 534–553 (1987). https://doi.org/10.1007/BF01937276
Received:
Revised:
Issue Date:
DOI: https://doi.org/10.1007/BF01937276