References & Citations
Mathematics > Numerical Analysis
Title: Efficient approximation of high-dimensional functions with neural networks
(Submitted on 9 Dec 2019 (v1), last revised 29 Jan 2021 (this version, v3))
Abstract: In this paper, we develop a framework for showing that neural networks can overcome the curse of dimensionality in different high-dimensional approximation problems. Our approach is based on the notion of a catalog network, which is a generalization of a standard neural network in which the nonlinear activation functions can vary from layer to layer as long as they are chosen from a predefined catalog of functions. As such, catalog networks constitute a rich family of continuous functions. We show that under appropriate conditions on the catalog, catalog networks can efficiently be approximated with rectified linear unit-type networks and provide precise estimates on the number of parameters needed for a given approximation accuracy. As special cases of the general results, we obtain different classes of functions that can be approximated with ReLU networks without the curse of dimensionality.
Submission history
From: Florian Rossmannek [view email][v1] Mon, 9 Dec 2019 19:00:14 GMT (23kb)
[v2] Thu, 1 Oct 2020 10:33:19 GMT (207kb,D)
[v3] Fri, 29 Jan 2021 07:55:05 GMT (275kb,D)
Link back to: arXiv, form interface, contact.