An Analysis of Stochastic Variance Reduced Gradient for Linear Inverse Problems

Abstract

Stochastic variance reduced gradient (SVRG) is a popular variance reduction technique for accelerating stochastic gradient descent (SGD). We provide a first analysis of the method for solving a class of linear inverse problems in the lens of the classical regularization theory. We prove that for a suitable constant step size schedule, the method can achieve an optimal convergence rate in terms of the noise level (under suitable regularity condition) and the variance of the SVRG iterate error is smaller than that by SGD. These theoretical findings are corroborated by a set of numerical experiments.

arXiv comment31 pages, 2 figures, to appear at Inverse Problems

References in corpus (8)

• SAGA: A Fast Incremental Gradient Method With Support for Non-Strongly Convex Composite Objectives
• Stochastic Variance Reduction for Nonconvex Optimization
• Stop Wasting My Gradients: Practical SVRG
• Optimal Rates for Multi-pass Stochastic Gradient Methods
• On the Regularizing Property of Stochastic Gradient Descent
• Online learning in optical tomography: a stochastic approach
• Optimal-order convergence of Nesterov acceleration for linear ill-posed problems
• Reconstructing the thermal phonon transmission coefficient at solid interfaces in the phonon transport equation