当前位置: X-MOL 学术arXiv.cs.LG › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Deep Neural Tangent Kernel and Laplace Kernel Have the Same RKHS
arXiv - CS - Machine Learning Pub Date : 2020-09-22 , DOI: arxiv-2009.10683
Lin Chen, Sheng Xu

We prove that the reproducing kernel Hilbert spaces (RKHS) of a deep neural tangent kernel and the Laplace kernel include the same set of functions, when both kernels are restricted to the sphere $\mathbb{S}^{d-1}$. Additionally, we prove that the exponential power kernel with a smaller power (making the kernel more non-smooth) leads to a larger RKHS, when it is restricted to the sphere $\mathbb{S}^{d-1}$ and when it is defined on the entire $\mathbb{R}^d$.

中文翻译:

Deep Neural Tangent Kernel 和 Laplace Kernel 具有相同的 RKHS

我们证明了深度神经切线核和拉普拉斯核的再生核希尔伯特空间 (RKHS) 包括相同的函数集,当两个核都被限制在球体 $\mathbb{S}^{d-1}$ 时。此外,我们证明了具有较小幂的指数幂内核(使内核更不光滑)导致更大的 RKHS,当它被限制在球体 $\mathbb{S}^{d-1}$ 和当它是在整个 $\mathbb{R}^d$ 上定义的。
更新日期:2020-10-02
down
wechat
bug