# NIPS Proceedingsβ

## Limitations on Variance-Reduction and Acceleration Schemes for Finite Sums Optimization

[PDF] [BibTeX] [Supplemental] [Reviews]

### Abstract

We study the conditions under which one is able to efficiently apply variance-reduction and acceleration schemes on finite sums problems. First, we show that perhaps surprisingly, the finite sum structure, by itself, is not sufficient for obtaining a complexity bound of $\tilde{\cO}((n+L/\mu)\ln(1/\epsilon))$ for $L$-smooth and $\mu$-strongly convex finite sums - one must also know exactly which individual function is being referred to by the oracle at each iteration. Next, we show that for a broad class of first-order and coordinate-descent finite sums algorithms (including, e.g., SDCA, SVRG, SAG), it is not possible to get an `accelerated' complexity bound of $\tilde{\cO}((n+\sqrt{n L/\mu})\ln(1/\epsilon))$, unless the strong convexity parameter is given explicitly. Lastly, we show that when this class of algorithms is used for minimizing $L$-smooth and non-strongly convex finite sums, the optimal complexity bound is $\tilde{\cO}(n+L/\epsilon)$, assuming that (on average) the same update rule is used for any iteration, and $\tilde{\cO}(n+\sqrt{nL/\epsilon})$, otherwise.