Optimal Approximation Rates for Deep ReLU Neural Networks on Sobolev and Besov Spaces

Open in new window