Lyapunov stability of the subgradient method with constant step size

DOI10.1007/S10107-023-01936-6MaRDI QIDQ6052064zbMATH OpenOpenAlexFDO

Authors Cédric Josz, Lexiao Lai

Publication date 23 October 2023

Published in Mathematical Programming. Series A. Series B (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2211.14850

zbMATH Keywords

differential inclusionsLyapunov stabilitysemi-algebraic geometry

Mathematics Subject Classification ID

Numerical mathematical programming methods (65K05) Nonlinear programming (90C30)

Abstract: We consider the subgradient method with constant step size for minimizing locally Lipschitz semi-algebraic functions. In order to analyze the behavior of its iterates in the vicinity of a local minimum, we introduce a notion of discrete Lyapunov stability and propose necessary and sufficient conditions for stability.

Recommendations

• Sufficient Conditions for Instability of the Subgradient Method with Constant Step Size
• Long term dynamics of the subgradient method for Lipschitz path differentiable functions
• Global stability of first-order methods for coercive tame functions
• A note on the convergence of subgradient optimization methods
• scientific article; zbMATH DE number 3995464

Cites work

• Clarke Subgradients of Stratifiable Functions
• Conservative set valued fields, automatic differentiation, stochastic gradient methods and deep learning
• Constant step stochastic approximations involving differential inclusions: stability, long-run convergence and applications
• Convergence of constant step stochastic gradient descent for non-smooth non-convex functions
• Convergence of the Iterates of Descent Methods for Analytic Cost Functions
• Convex optimization algorithms
• Curves of descent
• Generic Differentiability of Lipschitzian Functions
• Genericity in polynomial optimization
• Geometric categories and o-minimal structures
• scientific article; zbMATH DE number 3855514 (Why is no real title available?)
• scientific article; zbMATH DE number 4029737 (Why is no real title available?)
• scientific article; zbMATH DE number 46303 (Why is no real title available?)
• scientific article; zbMATH DE number 1160037 (Why is no real title available?)
• scientific article; zbMATH DE number 3805870 (Why is no real title available?)
• scientific article; zbMATH DE number 3187156 (Why is no real title available?)
• Low-rank matrix recovery with composite optimization: good conditioning and rapid convergence
• On gradients of functions definable in o-minimal structures
• On the division of distributions by polynomials
• On the stable equilibrium points of gradient systems
• Proximal Alternating Minimization and Projection Methods for Nonconvex Problems: An Approach Based on the Kurdyka-Łojasiewicz Inequality
• Stochastic Approximations and Differential Inclusions
• Stochastic subgradient method converges on tame functions
• Subgradient methods for sharp weakly convex functions
• Sur le problème de la division

Cited in

(8)

• Publisher correction to: ``Lyapunov stability of the subgradient method with constant step size
• Sufficient Conditions for Instability of the Subgradient Method with Constant Step Size
• A subgradient method with constant step-size for \(\ell_1\)-composite optimization
• Long term dynamics of the subgradient method for Lipschitz path differentiable functions
• Global stability of first-order methods for coercive tame functions
• Learning-rate-free momentum SGD with reshuffling converges in nonsmooth nonconvex optimization
• A minimization approach for minimax optimization with coupled constraints
• Inexact subgradient methods for semialgebraic functions

This page was built for publication: Lyapunov stability of the subgradient method with constant step size

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6052064)