1 paper
Patrick GelÃ, Aizhan Issagali, Ralf Kornhuber
We present a novel and mathematically transparent approach to function approximation and the training of large, high-dimensional neural networks, based on the approximate least-squ…