SGD without Replacement: Sharper Rates for General Smooth Convex Functions

Abstract

We study stochastic gradient descent {\em without replacement} (\sgdwor) for smooth convex functions. \sgdwor is widely observed to converge faster than true \sgd where each sample is drawn independently {\em with replacement} \cite{bottou2009curiously} and hence, is more popular in practice. But it's convergence properties are not well understood as sampling without replacement leads to coupling between iterates and gradients. By using method of exchangeable pairs to bound Wasserstein distance, we provide the first non-asymptotic results for \sgdwor when applied to {\em general smooth, strongly-convex} functions. In particular, we show that \sgdwor converges at a rate of $O(1/K^2)$ while \sgd is known to converge at $O(1/K)$ rate, where $K$ denotes the number of passes over data and is required to be {\em large enough}. Existing results for \sgdwor in this setting require additional {\em Hessian Lipschitz assumption} \cite{gurbuzbalaban2015random,haochen2018random}. For {\em small} $K$, we show \sgdwor can achieve same convergence rate as \sgd for {\em general smooth strongly-convex} functions. Existing results in this setting require $K=1$ and hold only for generalized linear models \cite{shamir2016without}. In addition, by careful analysis of the coupling, for both large and small $K$, we obtain better dependence on problem dependent parameters like condition number.

arXiv commentVersion 2 corrects a minor error in the bounds for K

References in corpus (2)

• Why Random Reshuffling Beats Stochastic Gradient Descent
• A note on the non-commutative arithmetic-geometric mean inequality

Cited by in corpus (9)

• Review: Deep Learning in Electron Microscopy
• Multiplicative noise and heavy tails in stochastic optimization
• A Unified Convergence Analysis for Shuffling-Type Gradient Methods
• Recht-Ré Noncommutative Arithmetic-Geometric Mean Conjecture is False
• An Empirical Study on the Intrinsic Privacy of SGD
• SMG: A Shuffling Gradient-Based Method with Momentum
• Permutation-Based SGD: Is Random Optimal?
• Convergence of Random Reshuffling Under The Kurdyka-Łojasiewicz Inequality
• Optimal Rates for Random Order Online Optimization