1: \begin{abstract}

2: We describe a novel optimization method for finite sums (such as empirical

3: risk minimization problems) building on the recently introduced SAGA

4: method. Our method achieves an accelerated convergence rate on strongly

5: convex smooth problems. Our method has only one parameter (a step

6: size), and is radically simpler than other accelerated methods for

7: finite sums. Additionally it can be applied when the terms are non-smooth,

8: yielding a method applicable in many areas where operator splitting

9: methods would traditionally be applied.

10: \end{abstract}

11: