Deep learning as optimal control problems and Riemannian discrete gradient descent.

To upload material, you must first log in.