DC Neural Networks avoid overfitting in one-dimensional nonlinear regression

Cesar Beltran-Royo, Laura Llopis-Ibor, Juan José Pantrigo, Iván Ramírez

Abstract: Highlights•We show the surprising ability of DC Neural Networks to avoid overfitting in nonlinear regression.•That is, DC Neural Networks self-regularize (do not require additional regularization techniques).•Thus, DC Neural Networks could result very useful for practical purposes based on nonlinear regression.