Sciweavers

NECO
2008
146views more  NECO 2008»
14 years 11 days ago
Deep, Narrow Sigmoid Belief Networks Are Universal Approximators
In this paper we show that exponentially deep belief networks [3, 7, 4] can approximate any distribution over binary vectors to arbitrary accuracy, even when the width of each lay...
Ilya Sutskever, Geoffrey E. Hinton