A Mean-Field Optimal Control Formulation of Deep Learning | Read Paper on Bytez