Spectral Barron space for deep neural network approximation
Fuente:
arXiv
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2023
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908449967702016 |
|---|---|
| author | Liao, Yulei Ming, Pingbing |
| author_facet | Liao, Yulei Ming, Pingbing |
| contents | We prove the sharp embedding between the spectral Barron space and the Besov space with embedding constants independent of the input dimension. Given the spectral Barron space as the target function space, we prove a dimension-free convergence result that if the neural network contains $L$ hidden layers with $N$ units per layer, then the upper and lower bounds of the $L^2$-approximation error are $\mathcal{O}(N^{-sL})$ with $0 < sL\le 1/2$, where $s\ge 0$ is the smoothness index of the spectral Barron space. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2309_00788 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Spectral Barron space for deep neural network approximation Liao, Yulei Ming, Pingbing Numerical Analysis 32C22, 32K05, 33C20, 41A25, 41A46, 42A38, 68T07 We prove the sharp embedding between the spectral Barron space and the Besov space with embedding constants independent of the input dimension. Given the spectral Barron space as the target function space, we prove a dimension-free convergence result that if the neural network contains $L$ hidden layers with $N$ units per layer, then the upper and lower bounds of the $L^2$-approximation error are $\mathcal{O}(N^{-sL})$ with $0 < sL\le 1/2$, where $s\ge 0$ is the smoothness index of the spectral Barron space. |
| title | Spectral Barron space for deep neural network approximation |
| topic | Numerical Analysis 32C22, 32K05, 33C20, 41A25, 41A46, 42A38, 68T07 |
| url | https://arxiv.org/abs/2309.00788 |