SVD-MPE: An SVD-Based Vector Extrapolation Method of Polynomial Type

An important problem that arises in different areas of science and engineering is that of computing the limits of sequences of vectors $\{\xx_m\}$, where $\xx_m\in \C^N$, $N$ being very large. Such sequences arise, for example, in the solution of systems of linear or nonlinear equations by fixed-point iterative methods, and $\lim_{m\to\infty}\xx_m$ are simply the required solutions. In most cases of interest, however, these sequences converge to their limits extremely slowly. One practical way to make the sequences $\{\xx_m\}$ converge more quickly is to apply to them vector extrapolation methods. Two types of methods exist in the literature: polynomial type methods and epsilon algorithms. In most applications, the polynomial type methods have proved to be superior convergence accelerators. Three polynomial type methods are known, and these are the {minimal polynomial extrapolation} (MPE), the {reduced rank extrapolation} (RRE), and the {modified minimal polynomial extrapolation} (MMPE). In this work, we develop yet another polynomial type method, which is based on the singular value decomposition, as well as the ideas that lead to MPE. We denote this new method by SVD-MPE. We also design a numerically stable algorithm for its implementation, whose computational cost and storage requirements are minimal. Finally, we illustrate the use of {SVD-MPE} with numerical examples.

[1]  Avram Sidi,et al.  Vector extrapolation methods with applications to solution of large systems of equations and to PageRank computations , 2008, Comput. Math. Appl..

[2]  S. Cabay,et al.  A Polynomial Extrapolation Method for Finding Limits and Antilimits of Vector Sequences , 1976 .

[3]  Khalide Jbilou,et al.  LU implementation of the modified minimal polynomial extrapolation method for solving linear and nonlinear systems , 1999 .

[4]  A. Sidi Extrapolation vs. projection methods for linear systems of equations , 1988 .

[5]  P. Wynn,et al.  On a device for computing the _{}(_{}) tranformation , 1956 .

[6]  Y. Saad,et al.  GMRES: a generalized minimal residual algorithm for solving nonsymmetric linear systems , 1986 .

[7]  M. Mesina Convergence acceleration for the iterative solution of the equations X = AX + f , 1977 .

[8]  Avram Sidi,et al.  Upper bounds for convergence rates of vector extrapolation methods on linear systems with initial iterations. Ph.D. Thesis , 1992 .

[9]  J. Stoer,et al.  Introduction to Numerical Analysis , 2002 .

[10]  D. Shanks Non‐linear Transformations of Divergent and Slowly Convergent Sequences , 1955 .

[11]  Extrapolation methods for vector sequences , 1992 .

[12]  P. Wynn,et al.  On a Device for Computing the e m (S n ) Transformation , 1956 .

[13]  W. Arnoldi The principle of minimized iterations in the solution of the matrix eigenvalue problem , 1951 .

[14]  Avram Sidi Convergence and stability properties of minimal polynomial and reduced rank extrapolation algorithms , 1986 .

[15]  C. Kelley Iterative Methods for Linear and Nonlinear Equations , 1987 .

[16]  Gene H. Golub,et al.  Matrix computations , 1983 .

[17]  Alston S. Householder,et al.  The Theory of Matrices in Numerical Analysis , 1964 .

[18]  Charles R. Johnson,et al.  Matrix analysis , 1985, Statistical Inference for Engineers and Data Scientists.

[19]  P. Wynn,et al.  Acceleration techniques for iterated vector and matrix problems : (mathematics of computation, _1_6(1962), nr 79, p 301-322) , 1962 .

[20]  Avram Sidi,et al.  Upper bounds for convergence rates of acceleration methods with initial iterations , 1998, Numerical Algorithms.

[21]  B. P. Pugachev Acceleration of the convergence of iterative processes and a method of solving systems of non-linear equations , 1977 .

[22]  Avram Sidi,et al.  Efficient implementation of minimal polynomial and reduced rank extrapolation methods , 1991 .

[23]  C. Lanczos Solution of Systems of Linear Equations by Minimized Iterations1 , 1952 .

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

[25]  P. R. Graves-Morris,et al.  Vector valued rational interpolants I. , 1983 .

[26]  P. Wynn General purpose vector epsilon algorithmAlgol procedures , 1964 .

[27]  Avram Sidi,et al.  Review of two vector extrapolation methods of polynomial type with applications to large-scale problems , 2012, J. Comput. Sci..

[28]  E. Gekeler,et al.  On the solution of systems of equations by the epsilon algorithm of Wynn , 1972 .

[29]  David A. Smith,et al.  Acceleration of convergence of vector sequences , 1986 .

[30]  C. Brezinski Generalisations de la transformation de shanks, de la table de Pade et de l'ε-algorithme , 1975 .

[31]  Philipp Birken,et al.  Numerical Linear Algebra , 2011, Encyclopedia of Parallel Computing.

[32]  P. Wynn,et al.  Continued fractions whose coefficients obey a non-commutative law of multiplication , 1963 .

[33]  Avram Sidi Convergence of intermediate rows of minimal polynomial and reduced rank extrapolation tables , 2005, Numerical Algorithms.

[34]  Tony F. Chan,et al.  An Improved Algorithm for Computing the Singular Value Decomposition , 1982, TOMS.

[35]  R. P. Eddy EXTRAPOLATING TO THE LIMIT OF A VECTOR SEQUENCE , 1979 .

[36]  J. Stein,et al.  Least-square acceleration of iterative methods for linear equations , 1974 .

[37]  Avram Sidi,et al.  Convergence and stability analyses for some vector extrapolation methods on the presence of defective iteration matrices , 1988 .