Regularization of limited memory quasi-Newton methods for large-scale nonconvex minimization

DOI10.1007/s12532-023-00238-4zbMath1522.90265arXiv1911.04584OpenAlexW2986149466MaRDI QIDQ6095733

Christian Kanzow, Daniel Steck

Publication date: 8 September 2023

Published in: Mathematical Programming Computation (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/1911.04584

zbMATH Keywords

global convergencelarge-scale optimizationquasi-Newton methodsL-BFGSregularized Newton methodslimited memory methods

Mathematics Subject Classification ID

Large-scale problems in mathematical programming (90C06) Nonconvex programming, global optimization (90C26) Numerical optimization and variational techniques (65K10) Methods of quasi-Newton type (90C53)

Cites Work

• Unnamed Item
• On solving L-SR1 trust-region subproblems
• Convergence properties of the regularized Newton method for the unconstrained nonconvex optimization
• A regularized Newton method without line search for unconstrained optimization
• On the limited memory BFGS method for large scale optimization
• Representations of quasi-Newton matrices and their use in limited memory methods
• A new regularized quasi-Newton method for unconstrained optimization
• A limited-memory multipoint symmetric secant method for bound constrained optimization
• Regularized Newton methods for convex minimization problems with singular solutions
• CUTEst: a constrained and unconstrained testing environment with safe threads for mathematical optimization
• On efficiently combining limited-memory and trust-region techniques
• New quasi-Newton methods for unconstrained optimization problems
• On the Global Convergence of the BFGS Method for Nonconvex Unconstrained Optimization Problems
• Shifted L-BFGS systems
• Statistical Quasi-Newton: A New Look at Least Change
• Testing a Class of Methods for Solving Minimization Problems with Simple Bounds on the Variables
• A Relationship between the BFGS and Conjugate Gradient Algorithms and Its Implications for New Algorithms
• Updating Quasi-Newton Matrices with Limited Storage
• Quasi-Newton Methods, Motivation and Theory
• Line search algorithms with guaranteed sufficient decrease
• Trust Region Methods
• Compact representation of the full Broyden class of quasi‐Newton updates
• Analysis of a Symmetric Rank-One Trust Region Method
• A New Algorithm for Unconstrained Optimization
• A modified BFGS method and its global convergence in nonconvex minimization
• Benchmarking optimization software with performance profiles.