B1384
Title: Approximation of functions from Korobov spaces by shallow neural networks
Authors: Yuqing Liu - City University of Hong Kong (Hong Kong) [presenting]
Abstract: The work is the first novel result of the approximability of shallow neural networks on Korobov spaces. A dimensional independent rate of approximating functions from the Korobov space by ReLU shallow neural networks will be given out. A careful Fourier analysis and the probability method often applied to get dimension-independent bounds will be explained in detail, after which one will find the approximation rate and estimation error bound to follow. Finally, an example will be provided as a justification for the sufficiency of the main result.
