A new algorithm for singular value decomposition and its parallelization

An algorithm mainly consisting of a part of Divide and Conquer and the twisted factorization is proposed for bidiagonal SVD. The algorithm costs O(n^2)flops and is highly parallelizable when singular values are isolated. If strong clusters exist, the singular vector computation needs reorthgonalization. In such case, the cost of the algorithm increases to O(n^2+nk^2)flops and the parallelism may worsen depending on the distribution of singular values. Here k is the size of the largest cluster. The algorithm needs only O(n) working memory for every type of matrices.

[1]  R Jessup,et al.  A Parallel Algorithm for Computing the Singular Value Decomposition of a Matrix:A Revision of Argonne National Laboratory Tech. Report ANL/MCS-TM-102 ; CU-CS-623-92 , 1994 .

[2]  Kinji Kimura,et al.  Recent Developments of the mdLVs Algorithm for Singular Values and the I-SVD Algorithm for Singular Value Decomposition (Fast Algorithms in Computational Fluids : theory and applications) , 2008 .

[3]  Robert A. van de Geijn,et al.  A Parallel Eigensolver for Dense Symmetric Matrices Based on Multiple Relatively Robust Representations , 2005, SIAM J. Sci. Comput..

[4]  William K. Pratt,et al.  Digital image processing (2nd ed.) , 1991 .

[5]  Kinji Kimura,et al.  Verification of dLVv Transformation for Singular Vector Computation with High Accuracy , 2006, PDPTA.

[6]  J. Cuppen.  A divide and conquer method for the symmetric tridiagonal eigenproblem , 1980 .

[7]  Yoshimasa Nakamura,et al.  Accurate computation of singular values in terms of shifted integrable schemes , 2006 .

[8]  Alston S. Householder,et al.  Handbook for Automatic Computation , 1960, Comput. J..

[9]  Inderjit S. Dhillon,et al.  Fernando's solution to Wilkinson's problem: An application of double factorization , 1997 .

[10]  S. Eisenstat,et al.  A Stable and Efficient Algorithm for the Rank-One Modification of the Symmetric Eigenproblem , 1994, SIAM J. Matrix Anal. Appl..

[11]  Jack J. Dongarra,et al.  A Parallel Divide and Conquer Algorithm for the Symmetric Eigenvalue Problem on Distributed Memory Architectures , 1999, SIAM J. Sci. Comput..

[12]  James Demmel,et al.  Applied Numerical Linear Algebra , 1997 .

[13]  Stanley C. Eisenstat,et al.  A Divide-and-Conquer Algorithm for the Symmetric Tridiagonal Eigenproblem , 1995, SIAM J. Matrix Anal. Appl..

[14]  K. V. Fernando.  On computing an eigenvector of a tridiagonal matrix , 1997 .

[15]  Jack Dongarra,et al.  ScaLAPACK Users' Guide , 1987 .

[16]  Kinji Kimura,et al.  An Evaluation of Singular Value Computation by the Discrete Lotka-Volterra System , 2005, PDPTA.

[17]  Inderjit S. Dhillon,et al.  Orthogonal Eigenvectors and Relative Gaps , 2003, SIAM J. Matrix Anal. Appl..

[18]  Masami Takata,et al.  A new singular value decomposition algorithm suited to parallelization and preliminary results , 2006 .

[19]  Stanley C. Eisenstat,et al.  A Divide-and-Conquer Algorithm for the Bidiagonal SVD , 1995, SIAM J. Matrix Anal. Appl..

[20]  James Demmel,et al.  Accurate Singular Values of Bidiagonal Matrices , 1990, SIAM J. Sci. Comput..

[21]  Gene H. Golub,et al.  Calculating the singular values and pseudo-inverse of a matrix , 2007, Milestones in Matrix Computation.