1 paper
L. Berlyand, O. Krupchytskyi, V. Slavin
This paper is concerned with a special class of deep neural networks (DNNs) where the input and the output vectors have the same dimension. Such DNNs are widely used in application…