Global Convergence of Stochastic Gradient Hamiltonian Monte Carlo for Nonconvex Stochastic Optimization: Nonasymptotic Performance Bounds and Momentum-Based Acceleration
Publication in refereed journal
Officially Accepted for Publication


Times Cited
Web of Science9WOS source URL (as at 19/08/2024) Click here for the latest count
Altmetrics Information
.

Other information
AbstractStochastic gradient Hamiltonian Monte Carlo (SGHMC) is a variant of stochastic gradients with momentum where a controlled and properly scaled Gaussian noise is added to the stochastic gradients to steer the iterates toward a global minimum. Many works report its empirical success in practice for solving stochastic nonconvex optimization problems; in particular, it has been observed to outperform overdamped Langevin Monte Carlo–based methods, such as stochastic gradient Langevin dynamics (SGLD), in many applications. Although the asymptotic global convergence properties of SGHMC are well known, its finite-time performance is not well understood. In this work, we study two variants of SGHMC based on two alternative discretizations of the underdamped Langevin diffusion. We provide finite-time performance bounds for the global convergence of both SGHMC variants for solving stochastic nonconvex optimization problems with explicit constants. Our results lead to nonasymptotic guarantees for both population and empirical risk minimization problems. For a fixed target accuracy level on a class of nonconvex problems, we obtain complexity bounds for SGHMC that can be tighter than those available for SGLD.
Acceptance Date06/04/2021
All Author(s) ListXuefeng Gao, Mert Gürbüzbalaban, Lingjiong Zhu
Journal nameOperations Research
Year2021
ISSN0030-364X
eISSN1526-5463
LanguagesEnglish-United States

Last updated on 2024-21-08 at 01:54