Abstract:We characterize the continuous isotropic positive definite kernels on $\mathbb{R}^d$, where isotropy refers to invariance under the orthogonal group $O(d)$ but not necessarily stationarity. Furthermore, we characterize strict positive definiteness for such kernels.
The class of isotropic kernels is fairly general as it unifies stationary isotropic and dot product kernels, and includes neural network kernels that arise from infinite-width limits of neural networks. As an application, we further characterize the continuous isotropic Gaussian random functions in terms of a series representation.
From: Felix Benning [view email]
[v1]
Fri, 27 Jun 2025 09:44:58 UTC (37 KB)
[v2]
Sun, 19 Jul 2026 09:22:11 UTC (48 KB)
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。