Saved in:
Bibliographic Details
Main Authors: Chen, Tyler, Huber, Caroline, Lin, Ethan, Zaid, Hajar
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2501.18717
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912883307184128
author Chen, Tyler
Huber, Caroline
Lin, Ethan
Zaid, Hajar
author_facet Chen, Tyler
Huber, Caroline
Lin, Ethan
Zaid, Hajar
contents We describe a randomized variant of the block conjugate gradient method for solving a single positive-definite linear system of equations. Our method provably outperforms preconditioned conjugate gradient with a broad-class of Nyström-based preconditioners, without ever explicitly constructing a preconditioner. In analyzing our algorithm, we derive theoretical guarantees for new variants of Nyström preconditioned conjugate gradient which may be of separate interest. We also describe how our approach yields state-of-the-art algorithms for key data-science tasks such as computing the entire ridge regression regularization path and generating multiple independent samples from a high-dimensional Gaussian distribution.
format Preprint
id arxiv_https___arxiv_org_abs_2501_18717
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Preconditioning without a preconditioner: faster ridge-regression and Gaussian sampling with randomized block Krylov subspace methods
Chen, Tyler
Huber, Caroline
Lin, Ethan
Zaid, Hajar
Numerical Analysis
We describe a randomized variant of the block conjugate gradient method for solving a single positive-definite linear system of equations. Our method provably outperforms preconditioned conjugate gradient with a broad-class of Nyström-based preconditioners, without ever explicitly constructing a preconditioner. In analyzing our algorithm, we derive theoretical guarantees for new variants of Nyström preconditioned conjugate gradient which may be of separate interest. We also describe how our approach yields state-of-the-art algorithms for key data-science tasks such as computing the entire ridge regression regularization path and generating multiple independent samples from a high-dimensional Gaussian distribution.
title Preconditioning without a preconditioner: faster ridge-regression and Gaussian sampling with randomized block Krylov subspace methods
topic Numerical Analysis
url https://arxiv.org/abs/2501.18717