Saved in:
Bibliographic Details
Main Authors: Wortsman, Arie, Loureiro, Bruno
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2510.04780
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911194034470912
author Wortsman, Arie
Loureiro, Bruno
author_facet Wortsman, Arie
Loureiro, Bruno
contents In this work, we investigate high-dimensional kernel ridge regression (KRR) on i.i.d. Gaussian data with anisotropic power-law covariance. This setting differs fundamentally from the classical source & capacity conditions for KRR, where power-law assumptions are typically imposed on the kernel eigen-spectrum itself. Our contributions are twofold. First, we derive an explicit characterization of the kernel spectrum for polynomial inner-product kernels, giving a precise description of how the kernel eigen-spectrum inherits the data decay. Second, we provide an asymptotic analysis of the excess risk in the high-dimensional regime for a particular kernel with this spectral behavior, showing that the sample complexity is governed by the effective dimension of the data rather than the ambient dimension. These results establish a fundamental advantage of learning with power-law anisotropic data over isotropic data. To our knowledge, this is the first rigorous treatment of non-linear KRR under power-law data.
format Preprint
id arxiv_https___arxiv_org_abs_2510_04780
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Kernel ridge regression under power-law data: spectrum and generalization
Wortsman, Arie
Loureiro, Bruno
Machine Learning
In this work, we investigate high-dimensional kernel ridge regression (KRR) on i.i.d. Gaussian data with anisotropic power-law covariance. This setting differs fundamentally from the classical source & capacity conditions for KRR, where power-law assumptions are typically imposed on the kernel eigen-spectrum itself. Our contributions are twofold. First, we derive an explicit characterization of the kernel spectrum for polynomial inner-product kernels, giving a precise description of how the kernel eigen-spectrum inherits the data decay. Second, we provide an asymptotic analysis of the excess risk in the high-dimensional regime for a particular kernel with this spectral behavior, showing that the sample complexity is governed by the effective dimension of the data rather than the ambient dimension. These results establish a fundamental advantage of learning with power-law anisotropic data over isotropic data. To our knowledge, this is the first rigorous treatment of non-linear KRR under power-law data.
title Kernel ridge regression under power-law data: spectrum and generalization
topic Machine Learning
url https://arxiv.org/abs/2510.04780