Geometric Rates of Approximation by Neural Networks
Abstrakt
Sets of functions for which errors in approximation by neural networks converge to zero geometrically fast are described.
Sets of functions for which errors in approximation by neural networks converge to zero geometrically fast are described.