We prove that for analytic functions in low dimension, the convergence rate of the deep neural network approximation is exponential.
We prove that for analytic functions in low dimension, the convergence rate of the deep neural network approximation is exponential.