Theory I: Deep networks and the curse of dimensionality

Tomaso Poggio, Qianli LiaoBulletin 2018 · DOI

Abstract

We review recent work characterizing the classes of functions for which deep learning can be exponentially better than shallow learning.Deep convolutional networks are a special case of these conditions, though weight sharing is not the main reason for their exponential advantage.
BibTeX
@article{poggio2018theory,
  author = {Tomaso Poggio and Qianli Liao},
  title = {{Theory I: Deep networks and the curse of dimensionality}},
  journal = {Bulletin},
  year = {2018},
  doi = {10.24425/bpas.2018.125924},
}

Loading related content...

Figure