On the approximation of functions by tanh neural networks


METADATA ONLY
Loading...

Date

2021-04

Publication Type

Report

ETH Bibliography

yes

Citations

Altmetric
METADATA ONLY

Data

Rights / License

Abstract

We derive bounds on the error, in high-order Sobolev norms, incurred in the approximation of Sobolev-regular as well as analytic functions by neural networks with the hyperbolic tangent activation function. These bounds provide explicit estimates on the approximation error with respect to the size of the neural networks. We show that tanh neural networks with only two hidden layers suffice to approximate functions at comparable or better rates than much deeper ReLU neural networks.

Publication status

published

Editor

Book title

Volume

2021-14

Pages / Article No.

Publisher

Seminar for Applied Mathematics, ETH Zurich

Event

Edition / version

Methods

Software

Geographic location

Date collected

Date created

Subject

Deep learning; Neural networks; Tanh; Function approximation

Organisational unit

03851 - Mishra, Siddhartha / Mishra, Siddhartha check_circle

Notes

Funding

Related publications and datasets