Who Breaks Early, Looses: Goal Oriented Training of Deep Neural Networks Based on Port Hamiltonian Dynamics.

Julian Burghoff, Marc Heinrich Monells, Hanno Gottschalk

Browse the full ICANN paper archive.