Abstract: Stochastic Gradient Descent (SGD) is the method of choice for large scale problems, most notably in deep learning. Recent studies target improving convergence and speed of the SGD algorithm.
Some results have been hidden because they may be inaccessible to you
Show inaccessible results