Super-Acceleration with Cyclical Step-sizes

Baptiste Goujaud · Damien Scieur · Aymeric Dieuleveut · Adrien Taylor · Fabian Pedregosa

[ Abstract ]
Wed 30 Mar 3:30 a.m. PDT — 5 a.m. PDT


We develop a convergence-rate analysis of momentum with cyclical step-sizes. We show that under some assumption on the spectral gap of Hessians in machine learning, cyclical step-sizes are provably faster than constant step-sizes. More precisely, we develop a convergence rate analysis for quadratic objectives that provides optimal parameters and shows that cyclical learning rates can improve upon traditional lower complexity bounds. We further propose a systematic approach to design optimal first order methods for quadratic minimization with a given spectral structure. Finally, we provide a local convergence rate analysis beyond quadratic minimization for the proposed methods and illustrate our findings through benchmarks on least squares and logistic regression problems.

Chat is not available.