A Generalized Newton Method for Subgradient Systems

DOI10.1287/MOOR.2022.1320MaRDI QIDQ6122563zbMATH OpenOpenAlexFDO

Authors Pham Duy Khanh, Boris S. Mordukhovich, Vo Thanh Phat

Publication date 1 March 2024

Published in Mathematics of Operations Research (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2009.10551

zbMATH Keywords

gradient and subgradient systemsmetric regularity and subregularityNewton methodsprox-regular functionsregularized least square problemssecond-order generalized differentiationsuperlinear convergencetilt stability in optimizationvariational analysis

Mathematics Subject Classification ID

Nonsmooth analysis (49J52) Set-valued and variational analysis (49J53) Newton-type methods (49M15) Nonlinear programming (90C30) Methods of quasi-Newton type (90C53)

Abstract: This paper proposes and develops a new Newton-type algorithm to solve subdifferential inclusions defined by subgradients of extended-real-valued prox-regular functions. The proposed algorithm is formulated in terms of the second-order subdifferential of such functions that enjoys extensive calculus rules and can be efficiently computed for broad classes of extended-real-valued functions. Based on this and on metric regularity and subregularity properties of subgradient mappings, we establish verifiable conditions ensuring well-posedness of the proposed algorithm and its local superlinear convergence. The obtained results are also new for the class of equations defined by continuously differentiable functions with Lipschitzian gradients (${calC}^{1,1}$ functions), which is the underlying case of our consideration. The developed algorithms for prox-regular functions and its extension to a structured class of composite functions are formulated in terms of proximal mappings and forward-backward envelopes. Besides numerous illustrative examples and comparison with known algorithms for ${calC}^{1,1}$ functions and generalized equations, the paper presents applications of the proposed algorithms to regularized least square problems arising in statistics, machine learning, and related disciplines.

Recommendations

• Generalized damped Newton algorithms in nonsmooth optimization via second-order subdifferentials
• A globally convergent proximal Newton-type method in nonsmooth convex optimization
• Generalized Newton algorithms for tilt-stable minimizers in nonsmooth optimization
• Proximal Newton-type methods for minimizing composite functions
• An inexact regularized proximal Newton-type method for nonconvex composite optimization problems

Cited in

(11)

• Globally convergent coderivative-based generalized Newton methods in nonsmooth optimization
• Generalized Newton Method with Positive Definite Regularization for Nonsmooth Optimization Problems with Nonisolated Solutions
• Variational and strong variational convexity in infinite-dimensional variational analysis
• A quadratically convergent semismooth Newton method for nonlinear semidefinite programming without generalized Jacobian regularity
• Local minimizers of nonconvex functions in Banach spaces via Moreau envelopes
• The least singular value function in variational analysis
• Coderivative-based semi-Newton method in nonsmooth difference programming
• Geometric characterizations of Lipschitz stability for convex optimization problems
• Second-order subdifferential optimality conditions in nonsmooth optimization
• Inexact proximal methods for weakly convex functions
• Tilt stability on Riemannian manifolds with application to convergence analysis of generalized Riemannian Newton method

This page was built for publication: A Generalized Newton Method for Subgradient Systems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6122563)