Stochastic Runge-Kutta methods and adaptive SGD-G2 stochastic gradient descent

no code implementations • 20 Feb 2020 • Imen Ayadi, Gabriel Turinici

The minimization of the loss function is of paramount importance in deep neural networks.