A General Analysis of Example-Selection for Stochastic Gradient Descent

Combining two principles we already discussed into one algorithm. • Principle: Write your learning task as an optimization problem and solve it with a scalable optimization algorithm.