We consider theoretical and computational issues related to operator regression and how to design deep neural networks (DNNs) that could represent accurately linear and nonlinear operators, mapping ...