Deep ReLU neural networks in high-dimensional approximation
Abstract
We study the computation complexity of deep ReLU (Rectified Linear Unit) neural networks for the approximation of functions from the H\"older-Zygmund space of mixed smoothness defined on the -dimensional unit cube when the dimension may be very large. The approximation error is measured in the norm of isotropic Sobolev space. For every function from the H\"older-Zygmund space of mixed smoothness, we explicitly construct a deep ReLU neural network having an output that approximates with a prescribed accuracy , and prove tight dimension-dependent upper and lower bounds of the computation complexity of this approximation, characterized as the size and the depth of this deep ReLU neural network, explicitly in and . The proof of these results are in particular, relied on the approximation by sparse-grid sampling recovery based on the Faber series.
Keywords
Cite
@article{arxiv.2007.08729,
title = {Deep ReLU neural networks in high-dimensional approximation},
author = {Dinh Dũng and Van Kien Nguyen},
journal= {arXiv preprint arXiv:2007.08729},
year = {2021}
}
Comments
5 figures