In this paper we introduce new bounds on the approximation of functions in deep networks and in doing so introduce some new deep network architectures for function approximation. These results give some theoretical insight into the success of autoencoders and ResNets.
@article{arxiv.1803.02956,
title = {Some Approximation Bounds for Deep Networks},
author = {Brendan McCane and Lech Szymanski},
journal= {arXiv preprint arXiv:1803.02956},
year = {2018}
}