Convergence of CG and GMRES on a tridiagonal Toeplitz linear system (Q2458227): Difference between revisions

From MaRDI portal
Set OpenAlex properties.
ReferenceBot (talk | contribs)
Changed an Item
 
Property / cites work
 
Property / cites work: Superlinear CG convergence for special right-hand sides / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q4858543 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q4348513 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Residual-Minimizing Krylov Subspace Methods for Stabilized Discretizations of Convection-Diffusion Equations / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q3258508 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q3671491 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q4348514 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Methods of conjugate gradients for solving linear systems / rank
 
Normal rank
Property / cites work
 
Property / cites work: Expressions and bounds for the GMRES residual / rank
 
Normal rank
Property / cites work
 
Property / cites work: Estimates for Some Computational Techniques in Linear Algebra / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q3527493 / rank
 
Normal rank
Property / cites work
 
Property / cites work: On Meinardus' examples for the conjugate gradient method / rank
 
Normal rank
Property / cites work
 
Property / cites work: Vandermonde matrices with Chebyshev nodes / rank
 
Normal rank
Property / cites work
 
Property / cites work: The rate of convergence of GMRES on a tridiagonal Toeplitz linear system / rank
 
Normal rank
Property / cites work
 
Property / cites work: Least Squares Residuals and Minimal Residual Methods / rank
 
Normal rank
Property / cites work
 
Property / cites work: Convergence of GMRES for Tridiagonal Toeplitz Matrices / rank
 
Normal rank
Property / cites work
 
Property / cites work: The worst-case GMRES for normal matrices / rank
 
Normal rank
Property / cites work
 
Property / cites work: On the worst-case convergence of MR and CG for symmetric positive definite tridiagonal Toeplitz matrices / rank
 
Normal rank
Property / cites work
 
Property / cites work: Über eine Verallgemeinerung einer Ungleichung von L.V. Kantorowitsch / rank
 
Normal rank
Property / cites work
 
Property / cites work: A note on conjugate gradient convergence / rank
 
Normal rank
Property / cites work
 
Property / cites work: A note on conjugate gradient convergence. II / rank
 
Normal rank
Property / cites work
 
Property / cites work: Solution of Sparse Indefinite Systems of Linear Equations / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q4414854 / rank
 
Normal rank
Property / cites work
 
Property / cites work: GMRES: A Generalized Minimal Residual Algorithm for Solving Nonsymmetric Linear Systems / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q4171531 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Complete stagnation of GMRES / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q3128879 / rank
 
Normal rank

Latest revision as of 12:13, 27 June 2024

scientific article
Language Label Description Also known as
English
Convergence of CG and GMRES on a tridiagonal Toeplitz linear system
scientific article

    Statements

    Convergence of CG and GMRES on a tridiagonal Toeplitz linear system (English)
    0 references
    0 references
    0 references
    31 October 2007
    0 references
    Error bounds, and sometimes exact bounds are constructed for the residual vector when conjugare gradient (CG) or generalized minimal residual (GMRES) iterative methods are applied to the solution of a tridiagonal normal (or Hermitian positive definite) Toeplitz system \(Ax=b\) for some special choices of the vectors \(b\). The method used rests upon the fact that bounds for the residual norm, and hence information about the convergence of the method, can be obtained from \[ \varepsilon_k=\min\{\| \text{diag}(g)V_{k+1,N}^T u\| _2/\| g\| _2\}. \] The minimum is taken over all vectors \(u\) whose first component is 1, \(g\) is a vector depending on the method and \(V_{k+1,N}\) is a \((k+1)\times N\) Vandermonde matrix whose nodes are the eigenvalues of \(A\). In the tridiagonal Toeplitz case, the eigenvalues can be expessed in terms of the zeros of Chebyshev polynomials of the second kind. All the explicit expressions for \(\varepsilon_k\) are given in terms of these Chebyshev nodes and the three Toeplitz parameters. The results for the particular right-hand sides \(b\) can be used to obtain estimates for general \(b\).
    0 references
    0 references
    conjugate gradient method
    0 references
    MINRES
    0 references
    Krylov subspace method
    0 references
    Vandermonde matrix
    0 references
    Chebyshev polynomial
    0 references
    generalized minimal residual (GMRES) iterative methods
    0 references
    convergence
    0 references
    tridiagonal normal Toeplitz system
    0 references
    Hermitian positive definite Toeplitz system
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references