Refinements of Universal Approximation Results for Deep Belief Networks and Restricted Boltzmann Machines

Open in new window