Spotlight Poster
Near-Optimal Approximation of Matrix Functions by the Lanczos Method
Noah Amsel · Tyler Chen · Anne Greenbaum · Cameron Musco · Christopher Musco
West Ballroom A-D #6900
Abstract:
Approximating the action of a matrix function $f(\vec{A})$ on a vector $\vec{b}$ is an increasingly important primitive in machine learning, data science, and statistics, with applications such as sampling high dimensional Gaussians, Gaussian process regression and Bayesian inference, principle component analysis, and approximating Hessian spectral densities.Over the past decade, a number of algorithms enjoying strong theoretical guarantees have been proposed for this task.Many of the most successful belong to a family of algorithms called Krylov subspace methods.Remarkably, a classic Krylov subspace method, called the Lanczos method for matrix functions (Lanczos-FA), frequently outperforms newer methods in practice. Our main result is a theoretical justification for this finding: we show that, for a natural class of rational functions, Lanczos-FA matches the error of the best possible Krylov subspace method up to a multiplicative approximation factor. The approximation factor depends on the degree of $f(x)$'s denominator and the condition number of $\vec{A}$, but not on the number of iterations $k$. Our result provides a strong justification for the excellent performance of Lanczos-FA, especially on functions that are well approximated by rationals, such as the matrix square root.
Live content is unavailable. Log in and register to view live content