arXiv subjects

Victor Y. Pan

Publications and source records attributed to Victor Y. Pan.

47 records · Page 3Linked to original sources

arXiv2014

Nearly Optimal Computations with Structured Matrices

We estimate the Boolean complexity of multiplication of structured matrices by a vector and the solution of nonsingular linear systems of equations with these matrices. We study four basic most popular classes, that is, Toeplitz, Hankel, Cauchy and Van-der-monde matrices, for which the cited computational problems are equivalent to the task of polynomial multiplication and division and polynomial and rational multipoint evaluation and interpolation. The Boolean cost estimates for the latter problems have been obtained by Kirrinnis in \cite{kirrinnis-joc-1998}, except for rational interpolation, which we supply now. All known Boolean cost estimates for these problems rely on using Kronecker product. This implies the $d$-fold precision increase for the $d$-th degree output, but we avoid such an increase by relying on distinct techniques based on employing FFT. Furthermore we simplify the analysis and make it more transparent by combining the representation of our tasks and algorithms in terms of both structured matrices and polynomials and rational functions. This also enables further extensions of our estimates to cover Trummer's important problem and computations with the popular classes of structured matrices that generalize the four cited basic matrix classes.

cs.SC↗

arXiv2014

Accelerated Approximation of the Complex Roots of a Univariate Polynomial (Extended Abstract)

Highly efficient and even nearly optimal algorithms have been developed for the classical problem of univariate polynomial root-finding (see, e.g., \cite{P95}, \cite{P02}, \cite{MNP13}, and the bibliography therein), but this is still an area of active research. By combining some powerful techniques developed in this area we devise new nearly optimal algorithms, whose substantial merit is their simplicity, important for the implementation.

cs.SC↗

arXiv2013

Supporting GENP with Random Multipliers

We prove that standard Gaussian random multipliers are expected to stabilize numerically both Gaussian elimination with no pivoting and block Gaussian elimination. Our tests show similar results where we applied circulant random multipliers instead of Gaussian ones.

math.NA↗

arXiv2013

New Structured Matrix Methods for Real and Complex Polynomial Root-finding

We combine the known methods for univariate polynomial root-finding and for computations in the Frobenius matrix algebra with our novel techniques to advance numerical solution of a univariate polynomial equation, and in particular numerical approximation of the real roots of a polynomial. Our analysis and experiments show efficiency of the resulting algorithms.

math.NA↗

arXiv2013

Transformations of Matrix Structures Work Again II

Matrices with the structures of Toeplitz, Hankel, Vandermonde and Cauchy types are omnipresent in modern computation. The four classes have distinct features, but in 1990 we showed that Vandermonde and Hankel multipliers transform all these structures into each other and proposed to employ this property to extend any successful algorithm that inverts matrices of one of these classes to inverting matrices of the three other classes. The power of this approach was widely recognized later, when novel numerically stable algorithms solved nonsingular Toeplitz linear systems of equations at first in quadratic and then in nearly linear (versus classical cubic) arithmetic time by means of transforming Toeplitz into Cauchy matrix structures and then linking it to the Hierarchical Semiseparable matrix structure, a specialization of matrix representations employed by the Fast Multipole Method. We first cover the method of structure transformation comprehensively, then analyze the latter algorithms for Toeplitz linear systems and extend them to approximate the products of Vandermonde and Cauchy matrices by a vector and the solutions of Vandermonde and Cauchy linear systems of equations where they are nonsingular and well conditioned. We decrease the arithmetic cost of the known algorithms from quadratic to nearly linear, and similarly for the computations with the matrices of a more general class having structures of Vandermonde and Cauchy types and for polynomial and rational evaluation and interpolation. We also accelerate a little further the known numerical approximation algorithms for a nonsingular Toeplitz or Toeplitz-like linear system by employing distinct transformations of matrix structures, and we briefly discuss some natural research challenges, particularly some promising applications of our techniques to high precision computations.

math.NA↗

arXiv2013

Estimating the Norms of Random Circulant and Toeplitz Matrices and Their Inverses

We estimate the norms of standard Gaussian random Toeplitz and circulant matrices and their inverses, mostly by means of combining some basic techniques of linear algebra. In the case of circulant matrices we obtain sharp probabilistic estimates, which show that these matrices are expected to be very well conditioned. Our probabilistic estimates for the norms of standard Gaussian random Toeplitz matrices are within a factor of 1.4143 from those in the circulant case. We also achieve partial progress in estimating the norms of Toeplitz inverses. Namely we yield reasonable probabilistic upper estimates for these norms assuming certain bounds on the absolute values of two corner entries of the inverse. Empirically we observe that the condition numbers of random Toeplitz and general matrices tend to be of the same order. As the matrix size grows, these numbers grow equally slowly, although faster than for random circulant matrices.

math.NA↗

arXiv2013

Chapter 10: Algebraic Algorithms

Our Chapter in the upcoming Volume I: Computer Science and Software Engineering of Computing Handbook (Third edition), Allen Tucker, Teo Gonzales and Jorge L. Diaz-Herrera, editors, covers Algebraic Algorithms, both symbolic and numerical, for matrix computations and root-finding for polynomials and systems of polynomials equations. We cover part of these large subjects and include basic bibliography for further study. To meet space limitation we cite books, surveys, and comprehensive articles with pointers to further references, rather than including all the original technical papers.

cs.DS↗

arXiv2013

Transformations of Matrix Structures Work Again

In 1989 we proposed to employ Vandermonde and Hankel multipliers to transform into each other the matrix structures of Toeplitz, Hankel, Vandermonde and Cauchy types as a means of extending any successful algorithm for the inversion of matrices having one of these structures to inverting the matrices with the structures of the three other types. Surprising power of this approach has been demonstrated in a number of works, which culminated in ingeneous numerically stable algorithms that approximated the solution of a nonsingular Toeplitz linear system in nearly linear (versus previuosly cubic) arithmetic time. We first revisit this powerful method, covering it comprehensively, and then specialize it to yield a similar acceleration of the known algorithms for computations with matrices having structures of Vandermonde or Cauchy types. In particular we arrive at numerically stable approximate multipoint polynomial evaluation and interpolation in nearly linear arithmetic time.

math.NA↗

arXiv2012

More on the Power of Randomized Matrix Computations

A random matrix is likely to be well conditioned, and motivated by this well known property we employ random matrix multipliers to advance some fundamental matrix computations. This includes numerical stabilization of Gaussian elimination with no pivoting as well as block Gaussian elimination, approximation of the leading and trailing singular spaces of an ill conditioned matrix, associated with its largest and smallest singular values, respectively, and approximation of this matrix by low-rank matrices, with further extensions to computing numerical ranks and the approximation of tensor decomposition. We formally support the efficiency of the proposed techniques where we employ Gaussian random multipliers, but our extensive tests have consistently produced the same outcome where instead we used sparse and structured random multipliers, defined by much fewer random parameters compared to the number of their entries.

math.NA↗

arXiv2012

Condition Numbers of Random Toeplitz and Circulant Matrices

Estimating the condition numbers of random structured matrices is a well known challenge, linked to the design of efficient randomized matrix algorithms. We deduce such estimates for Gaussian random Toeplitz and circulant matrices. The former estimates can be surprising because the condition numbers grow exponentially in n as n grows to infinity for some large and important classes of n-by-n Toeplitz matrices, whereas we prove the opposit for Gaussian random Toeplitz matrices. Our formal estimates are in good accordance with our numerical tests, except that circulant matrices tend to be even better conditioned according to the tests than according to our formal study.

math.NA↗

arXiv2012

Randomized Matrix Computations

Random matrices tend to be well conditioned, and we employ this well known property to advance matrix computations. We prove that our algorithms employing Gaussian random matrices are efficient, but in our tests the algorithms have consistently remained as powerful where we used sparse and structured random matrices, defined by much fewer random parameters. We numerically stabilize Gaussian elimination with no pivoting as well as block Gaussian elimination, precondition an ill conditioned linear system of equations, compute numerical rank of a matrix without orthogonalization and pivoting, approximate the singular spaces of an ill conditioned matrix associated with its largest and smallest singular values, and approximate this matrix with low-rank matrices, with applications to its 2-by-2 block triangulation and to tensor decomposition. Some of our results and techniques can be of independent interest, e.g., our estimates for the condition numbers of random Toeplitz and circulant matrices and our variations of the Sherman--Morrison--Woodbury formula.

math.NA↗