Deep Network Approximation: Beyond ReLU to Diverse Activation Functions
This paper explores the expressive power of deep neural networks for a diverse range of activation functions. An activation function set π is defined to encompass the majority of commonly used activation functions, such as πππ»π, π»ππππ’πππ»π, πππ»π^2, π΄π»π, ππ΄π»π, ππππππππ, πΆπ΄π»π, πππ»π, ππ πππ, πΌπππ, πππππππ, ππππ, π°πππππ, ππππππππ, ππππ»π, and πππ. We demonstrate that for any activation function Ο±βπ, a πππ»π network of width N and depth L can be approximated to arbitrary precision by a Ο±-activated network of width 6N and depth 2L on any bounded set. This finding enables the extension of most approximation results achieved with πππ»π networks to a wide variety of other activation functions, at the cost of slightly larger constants.
READ FULL TEXT