Numerical Algorithms

, Volume 13, Issue 1, pp 123–152

Approximating dominant singular triplets of large sparse matrices via modified moments

  • Sowmini Varadhan
  • Michael W. Berry
  • Gene H. Golub
Article

Abstract

A procedure for determining a few of the largest singular values and corresponding singular vectors of large sparse matrices is presented. Equivalent eigensystems are solved using a technique originally proposed by Golub and Kent based on the computation of modified moments. The asynchronicity in the computations of moments and eigenvalues makes this method attractive for parallel implementations on a network of workstations. Although no obvious relationship between modified moments and the corresponding eigenvectors is known to exist, a scheme to approximate both eigenvalues and eigenvectors (and subsequently singular values and singular vectors) has been produced. This scheme exploits both modified moments in conjunction with the Chebyshev semi-iterative method and deflation techniques to produce approximate eigenpairs of the equivalent sparse eigensystems. The performance of an ANSI-C implementation of this scheme on a network of UNIX workstations and a 256-processor Cray T3D is presented.

Keywords

Chebyshev semi-iterative method modified moments singular value decomposition sparse matrices 

AMS subject classification

65K15 65K50 68P20 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. [1]
    R. Barrett et al.,Templastes for the Solution of Linear Systems: Building Blocks for Iterative Methods (SIAM, Philadelphia, 1994).Google Scholar
  2. [2]
    M. Berry and G. Golub, Estimating the largest singular values of large sparse matrices via modified moments. Numerical Algorithms 1 (1991) 353–374.Google Scholar
  3. [3]
    M. W. Berry, Large scale singular value computations, International Journal of Supercomputer Applications 6 (1992) 13–49.Google Scholar
  4. [4]
    M. W. Berry, S. T. Dumais and G. W. O'Brien, Using linear algebra for intelligent information retrieval, SIAM Review 37 (1995) 573–595.Google Scholar
  5. [5]
    C. Brezinski,Padé-type Approximation and General Orthogonal Polynomials (Birkhäuser, 1980).Google Scholar
  6. [6]
    Inc. Cray Research,UNICOS C Library Reference Manual (Cray Research, Inc., Eagan, MN, 5.0 ed., 1994).Google Scholar
  7. [7]
    J. Dongarra, J. Bunch, C. Moler and G. W. Stewart,Linpack Users' Guide (SIAM, Philadelphia, 1979).Google Scholar
  8. [8]
    W. Gautschi, On generating orthogonal polynomials, SIAM Journal of Statistical and Scientific Computing 3 (1982) 289–317.Google Scholar
  9. [9]
    A. Geist, A. Beguelin, J. Dongarra, W. Jiang, R. Manchek and V. Sunderam,PVM: Parallel Virtual Machine (MIT Press, Cambridge, MA, 1994).Google Scholar
  10. [10]
    G. Golub and M. D. Kent, Estimating eigenvalues for iterative methods, Mathematics of Computation 53 (1989) 619–626.Google Scholar
  11. [11]
    G. Golub and C. Van Loan,Matrix Computations (Johns Hopkins University Press, Baltimore, 2nd ed., 1989).Google Scholar
  12. [12]
    G. 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 Transactions on Mathematical Software 7 (1981) 149–169.Google Scholar
  13. [13]
    G. Golub and R. S. Varga, Chebyshev semi-iterative methods, successive overrelaxation iterative methods, and second order Richardson iterative methods, Numerische Mathematik 3 (1961) 147–156.Google Scholar
  14. [14]
    G. H. Golub, Some uses of the Lanczos algorithm in numerical linear algebra, in:Topics in Numerical Analysis, ed. J. Miller (1974).Google Scholar
  15. [15]
    A. S. Householder,The Theory of Matrices in Numerical Analysis (Blaisdell, New York, 1964).Google Scholar
  16. [16]
    C. Lanczos, An iteration method for the solution of the eigenvalue problem, J. Res. Nat. Bur. Standards 45 (1950) 255–282.Google Scholar
  17. [17]
    R. B. Lehoucq, D. C. Sorensen and P. Vu, ARPACK: An implementation of the Implicitly Restarted Arnoldi Iteration that computes some of the eigenvalues and eigenvectors of a large sparse matrix (1994); available from netlib@ornl.gov under the directory Scalapack.Google Scholar
  18. [18]
    C. C. Paige and M. A. Saunders, Solution of sparse indefinite systems of linear equations, SIAM Journal of Numerical Analysis 12 (1975), 617–629.Google Scholar
  19. [19]
    B. N. Parlett,The Symmetric Eigenvalue Problem (Prentice-Hall, Englewood Cliffs, NJ, 1980).Google Scholar
  20. [20]
    Y. Saad,Numerical Methods for Large Eigenvalue Problems (Manchester University Press, Manchester, UK, 1992).Google Scholar
  21. [21]
    G. W. StewartIntroduction to Matrix Computations (Academic Press, New York, 1973).Google Scholar
  22. [22]
    S. Varadhan, Estimating the largest singular values/vectors of large sparse matrices via modified moments, Technical Report CS-96-319, University of Tennessee, Knoxville, TN (February 1996).Google Scholar
  23. [23]
    R. S. Varga,Matrix Iterative Analysis (Prentice-Hall, Englewood Cliffs, NJ, 1962).Google Scholar

Copyright information

© J.C. Baltzer AG, Science Publishers 1996

Authors and Affiliations

  • Sowmini Varadhan
    • 1
  • Michael W. Berry
    • 1
  • Gene H. Golub
    • 2
  1. 1.Department of Computer ScienceUniversity of TennesseeKnoxvilleUSA
  2. 2.Department of Computer ScienceStanford UniversityStanfordUSA

Personalised recommendations