Title:  Mini-batch stochastic gradient descent with dynamic sample and step sizes

In this presentation we will look at stochastic gradient descent and its variants, with a particular focus on the mini-batch implementation. Basic convergence results will be presented as well as an overview of the current research done on dynamically chosen sample sizes. New sample and step size rules will be presented as well as a preliminary study examining the usefulness of the mini-batch methodology in an industrial application.