Optimal near-optimality bounds for the Lanczos method for matrix functions

DOI10.48550/arXiv.2608.07160MaRDI QIDQ7284682arXivFDO

Authors Tyler Chen, David Persson

Publication date 7 August 2026

Published in arXiv

Tldr We provide a new analysis of Lanczos approximation to matrix functions and show that for Stieltjes matrix functions, the error bound improves over previously known bounds by up to a factor of two. The result holds under mild assumptions on the function and has implications for linear algebra problems such as quadratic forms and matrix sign function computations.

Full work available at URL https://doi.org/10.48550/arXiv.2608.07160

Mathematics Subject Classification ID

Computational methods for sparse matrices (65F50) Numerical computation of matrix exponential and similar matrix functions (65F60) Analysis of algorithms and problem complexity (68Q25)

arXiv classification

math.NA

Cites work

• A numerical solution using an adaptively preconditioned Lanczos method for a class of linear systems related with the fractional Poisson equation
• A restarted Lanczos approximation to functions of a symmetric matrix
• Analysis of Some Krylov Subspace Approximations to the Matrix Exponential Operator
• Approximating the exponential, the lanczos method and an Õ(m)-time spectral algorithm for balanced separator
• Comparison of splittings used with the conjugate gradient algorithm
• Efficient and stable Arnoldi restarts for matrix functions based on quadrature
• Error Bounds for Lanczos-Based Matrix Function Approximation
• Error bounds in the simple Lanczos procedure for computing functions of symmetric matrices and eigenvalues
• Exponential integrators
• Fast estimation of \(\mathrm{tr}(f(A))\) via stochastic Lanczos quadrature
• Faster algorithms via approximation theory
• Krylov subspace approximation of eigenpairs and matrix functions in exact and computer arithmetic
• Matrices, moments and quadrature with applications
• Matrix functions
• Methods of conjugate gradients for solving linear systems
• Near instance optimality of the Lanczos method for Stieltjes and related matrix functions
• Nearly Optimal Approximation of Matrix Functions by the Lanczos Method
• Numerical methods for the QCDd overlap operator. I: Sign-function and error bounds
• On Krylov Subspace Approximations to the Matrix Exponential Operator
• Preconditioned Krylov subspace methods for sampling multivariate Gaussian distributions
• Stability of the Lanczos method for matrix function approximation
• The Lanczos algorithm for matrix functions: a handbook for scientists
• Using Nonorthogonal Lanczos Vectors in the Computation of Matrix Functions

This page was built for publication: Optimal near-optimality bounds for the Lanczos method for matrix functions

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q7284682)