Learning Functions: When Is Deep Better Than Shallow [article]

Hrushikesh Mhaskar, Qianli Liao, Tomaso Poggio
2016 arXiv   pre-print
While the universal approximation property holds both for hierarchical and shallow networks, we prove that deep (hierarchical) networks can approximate the class of compositional functions with the same accuracy as shallow networks but with exponentially lower number of training parameters as well as VC-dimension. This theorem settles an old conjecture by Bengio on the role of depth in networks. We then define a general class of scalable, shift-invariant algorithms to show a simple and natural
more » ... et of requirements that justify deep convolutional networks.
arXiv:1603.00988v4 fatcat:o5w4pcmyhfdkfmzbyq5ly37dgu