Enregistré dans:
Détails bibliographiques
Auteurs principaux: Shah, Rikhav, Silwal, Sandeep, Xu, Haike
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:https://arxiv.org/abs/2510.02540
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866914367020204032
author Shah, Rikhav
Silwal, Sandeep
Xu, Haike
author_facet Shah, Rikhav
Silwal, Sandeep
Xu, Haike
contents This paper studies the use of kernel density estimation (KDE) for linear algebraic tasks involving the kernel matrix of a collection of $n$ data points in $\mathbb R^d$. In particular, we improve upon existing algorithms for computing the following up to $(1+\varepsilon)$ relative error: matrix-vector products, matrix-matrix products, the spectral norm, and sum of all entries. The runtimes of our algorithms depend on the dimension $d$, the number of points $n$, and the target error $\varepsilon$. Importantly, the dependence on $n$ in each case is far lower when accessing the kernel matrix through KDE queries as opposed to reading individual entries. Our improvements over existing best algorithms (particularly those of Backurs, Indyk, Musco, and Wagner '21) for these tasks reduce the polynomial dependence on $\varepsilon$, and additionally decreases the dependence on $n$ in the case of computing the sum of all entries of the kernel matrix. We complement our upper bounds with several lower bounds for related problems, which provide (conditional) quadratic time hardness results and additionally hint at the limits of KDE based approaches for the problems we study.
format Preprint
id arxiv_https___arxiv_org_abs_2510_02540
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Even Faster Kernel Matrix Linear Algebra via Density Estimation
Shah, Rikhav
Silwal, Sandeep
Xu, Haike
Data Structures and Algorithms
Machine Learning
Numerical Analysis
68W25, 15B48, 15B05, 15A18
E.1; F.2.1
This paper studies the use of kernel density estimation (KDE) for linear algebraic tasks involving the kernel matrix of a collection of $n$ data points in $\mathbb R^d$. In particular, we improve upon existing algorithms for computing the following up to $(1+\varepsilon)$ relative error: matrix-vector products, matrix-matrix products, the spectral norm, and sum of all entries. The runtimes of our algorithms depend on the dimension $d$, the number of points $n$, and the target error $\varepsilon$. Importantly, the dependence on $n$ in each case is far lower when accessing the kernel matrix through KDE queries as opposed to reading individual entries. Our improvements over existing best algorithms (particularly those of Backurs, Indyk, Musco, and Wagner '21) for these tasks reduce the polynomial dependence on $\varepsilon$, and additionally decreases the dependence on $n$ in the case of computing the sum of all entries of the kernel matrix. We complement our upper bounds with several lower bounds for related problems, which provide (conditional) quadratic time hardness results and additionally hint at the limits of KDE based approaches for the problems we study.
title Even Faster Kernel Matrix Linear Algebra via Density Estimation
topic Data Structures and Algorithms
Machine Learning
Numerical Analysis
68W25, 15B48, 15B05, 15A18
E.1; F.2.1
url https://arxiv.org/abs/2510.02540