Matrices, Moments and Quadrature with Applications by Gene H. Golub

By Gene H. Golub

This computationally orientated publication describes and explains the mathematical relationships between matrices, moments, orthogonal polynomials, quadrature principles, and the Lanczos and conjugate gradient algorithms. The e-book bridges various mathematical components to acquire algorithms to estimate bilinear kinds concerning vectors and a functionality of the matrix. the 1st a part of the booklet offers the mandatory mathematical history and explains the speculation. the second one half describes the purposes and provides numerical examples of the algorithms and methods constructed within the first part.Applications addressed within the ebook contain computing components of services of matrices; acquiring estimates of the mistake norm in iterative equipment for fixing linear structures and computing parameters in least squares and overall least squares; and fixing ill-posed difficulties utilizing Tikhonov regularization.This ebook will curiosity researchers in numerical linear algebra and matrix computations, in addition to scientists and engineers engaged on difficulties regarding computation of bilinear kinds.

Show description

Read or Download Matrices, Moments and Quadrature with Applications (Princeton Series in Applied Mathematics) PDF

Similar linear books

Mengentheoretische Topologie

Eine verständliche und vollständige Einführung in die Mengentheoretische Topologie, die als Begleittext zu einer Vorlesung, aber auch zum Selbststudium für Studenten ab dem three. Semester bestens geeignet ist. Zahlreiche Aufgaben ermöglichen ein systematisches Erlernen des Stoffes, wobei Lösungshinweise bzw.

Combinatorial and Graph-Theoretical Problems in Linear Algebra

This IMA quantity in arithmetic and its functions COMBINATORIAL AND GRAPH-THEORETICAL difficulties IN LINEAR ALGEBRA is predicated at the lawsuits of a workshop that was once an essential component of the 1991-92 IMA software on "Applied Linear Algebra. " we're thankful to Richard Brualdi, George Cybenko, Alan George, Gene Golub, Mitchell Luskin, and Paul Van Dooren for making plans and enforcing the year-long software.

Linear Algebra and Matrix Theory

This revision of a well known textual content comprises extra subtle mathematical fabric. a brand new part on functions offers an creation to the fashionable remedy of calculus of numerous variables, and the idea that of duality gets accelerated assurance. Notations were replaced to correspond to extra present utilization.

Extra resources for Matrices, Moments and Quadrature with Applications (Princeton Series in Applied Mathematics)

Example text

Using the orthogonality of the Xi ’s and the spectral decomposition of A, we can write i δij I2 = XiT Xj = j (i) (Ck )T X0T Ak (j) Al X0 Cl k=0 l=0 (i) T (j) T k+l T (Ck ) X0 QΛ Q X0 Cl = k,l (i) ˆ k+l X ˆ T C (j) (Ck )T XΛ l = k,l n (i) ˆ ˆT λk+l m Xm X m (Ck )T = m=1 k,l n (i) λkm (Ck )T = m=1 (j) Cl (j) T ˆmX ˆm X k λlm Cl , l ˆ m are the columns of X ˆ = X T Q, which is a 2 × n matrix. Therefore, where X 0 n T ˆmX ˆm pi (λm )T X pj (λm ). XiT Xj = m=1 The sum in the right-hand side can be written as an integral for a 2 × 2 matrix measure,  if λ < λ1 ,  0, i ˆ ˆT α(λ) = j=1 Xj Xj , if λi ≤ λ < λi+1 ,  n ˆ ˆT Xj X , if λn ≤ λ.

The last sum can be written as an integral for a measure α which is piecewise constant (here we suppose for the sake of simplicity that the eigenvalues of A are distinct):  if λ < λ1 ,  0, i 2 [ˆ v ] , if λi ≤ λ < λi+1 , j α(λ) = j=1  n 2 [ˆ v ] , if λn ≤ λ. j=1 j THE LANCZOS AND CONJUGATE GRADIENT ALGORITHMS 43 The measure α has a finite number of points of increase at the (unknown) eigenvalues of A. ✷ The Lanczos algorithm provides an example of orthonormal polynomials for an unknown measure.

PN −1 (λ)]JN + [0, . . , 0, pN (λ)ΓN ], and as P (λ) = [p0 (λ), . . , pN −1 (λ)]T , JN P (λ) = λP (λ) − [0, . . 16). As in the scalar case, the inner product of the matrices Xi can be related to an integral. 6 Considering the matrices Xk , there exists a matrix measure α (defined in the proof) such that b XiT Xj = pi (λ)T dα(λ)pj (λ) = δij I2 , a where a ≤ λ1 = λmin and b ≥ λn = λmax . Proof. Using the orthogonality of the Xi ’s and the spectral decomposition of A, we can write i δij I2 = XiT Xj = j (i) (Ck )T X0T Ak (j) Al X0 Cl k=0 l=0 (i) T (j) T k+l T (Ck ) X0 QΛ Q X0 Cl = k,l (i) ˆ k+l X ˆ T C (j) (Ck )T XΛ l = k,l n (i) ˆ ˆT λk+l m Xm X m (Ck )T = m=1 k,l n (i) λkm (Ck )T = m=1 (j) Cl (j) T ˆmX ˆm X k λlm Cl , l ˆ m are the columns of X ˆ = X T Q, which is a 2 × n matrix.

Download PDF sample

Rated 4.71 of 5 – based on 20 votes