Data Sampling Strategies in Stochastic Algorithms for Empirical Risk Minimization

Dominik Csiba

Gradient descent methods and especially their stochastic variants have become highly popular in the last decade due to their efficiency on big data optimization problems. In this t…