Deep Neural Network Learning with Second-Order Optimizers -- a Practical Study with a Stochastic Quasi-Gauss-Newton Method

Training in supervised deep learning is computationally demanding, and the convergence behavior is usually not fully understood. We introduce and study a second-order stochastic quasi-Gauss--Newton (SQGN) optimization method that combines ideas from stochastic quasi-Newton methods, Gauss--Newton methods, and variance reduction to address this problem. SQGN provides excellent accuracy without the need for experimenting with many hyper-parameter configurations, which is often computationally prohibitive given the number of combinations and the cost of each training process. We discuss the implementation of SQGN with TensorFlow, and we compare its convergence and computational performance to selected first-order methods using the MNIST benchmark and a large-scale seismic tomography application from Earth science.