Anderson Acceleration of Proximal Gradient Methods

Abstract

Anderson acceleration is a well-established and simple technique for speeding up fixed-point computations with countless applications. Previous studies of Anderson acceleration in optimization have only been able to provide convergence guarantees for unconstrained and smooth problems. This work introduces novel methods for adapting Anderson acceleration to (non-smooth and constrained) proximal gradient algorithms. Under some technical conditions, we extend the existing local convergence results of Anderson acceleration for smooth fixed-point mappings to the proposed scheme. We also prove analytically that it is not, in general, possible to guarantee global convergence of native Anderson acceleration. We therefore propose a simple scheme for stabilization that combines the global worst-case guarantees of proximal gradient methods with the local adaptation and practical speed-up of Anderson acceleration.

arXiv comment25 pages, 7 figures

Cited by in corpus (6)

• Stochastic Anderson Mixing for Nonconvex Stochastic Optimization
• Weighted Low-Rank Matrix Approximation: Acceleration and Applications
• Nonmonotone Globalization for Anderson Acceleration via Adaptive Regularization
• A trust region-type normal map-based semismooth Newton method for nonsmooth nonconvex composite optimization
• Anderson acceleration of coordinate descent
• Anderson Acceleration Based on the ${\mathcal{H}}^{-s}$ Sobolev Norm for Contractive and Noncontractive Fixed-Point Operators