Contributed Video: DDPNOpt: Differential Dynamic Programming Neural Optimizer, Guan-Horng Liu

Interpretation of Deep Neural Networks (DNNs) training as an optimal control problem with nonlinear dynamical systems has received considerable attention recently, yet the algorithmic development remains relatively limited. In this work, we make an attempt along this line by first showing that most widely-used algorithms for training DNNs can be linked to the Differential Dynamic Programming (DDP), a celebrated second-order method rooted in trajectory optimization. In this vein, we propose a new class of optimizer, DDP Neural Optimizer (DDPNOpt), for training DNNs. DDPNOpt features layer-wise feedback policies which improve convergence and robustness. It outperforms other optimal-control inspired training methods in both convergence and complexity, and is competitive against state-of-the-art first and second order methods. Our work opens up new avenues for principled algorithmic design built upon the optimal control theory.