On solving large-scale limited-memory quasi-Newton equations

DOI10.1016/J.LAA.2016.11.003MaRDI QIDQ501277zbMATH OpenOpenAlexFDO

Authors Jennifer B. Erway, Roummel F. Marcia

Publication date 29 December 2016

Published in Linear Algebra and its Applications (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1510.06378

zbMATH Keywords

Sherman-Morrison-Woodbury formulalimited-memory quasi-Newton methodssymmetric rank-one updatecompact representationBroyden class of updatesBroyden-Fletcher-Goldfarb-Shanno updateDavidon-Fletcher-Powell update

Mathematics Subject Classification ID

Numerical computation of eigenvalues and eigenvectors of matrices (65F15) Iterative numerical methods for linear systems (65F10) Numerical computation of matrix norms, conditioning, scaling (65F35)

Abstract: We consider the problem of solving linear systems of equations arising with limited-memory members of the restricted Broyden class of updates and the symmetric rank-one (SR1) update. In this paper, we propose a new approach based on a practical implementation of the compact representation for the inverse of these limited-memory matrices. Numerical results suggest that the proposed method compares favorably in speed and accuracy to other algorithms and is competitive with several update-specific methods available to only a few members of the Broyden class of updates. Using the proposed approach has an additional benefit: The condition number of the system matrix can be computed efficiently.

Recommendations

• scientific article; zbMATH DE number 940915
• Limited-memory BFGS systems with diagonal updates
• scientific article; zbMATH DE number 1292995
• A compact limited memory method for large scale unconstrained optimization
• A limited memory method for large scale problems

Cites work

• scientific article; zbMATH DE number 5359577 (Why is no real title available?)
• scientific article; zbMATH DE number 5060482 (Why is no real title available?)
• scientific article; zbMATH DE number 961607 (Why is no real title available?)
• A Broyden class of quasi-Newton methods for Riemannian optimization
• Automatic Preconditioning by Limited Memory Quasi-Newton Updating
• Limited-memory BFGS systems with diagonal updates
• On efficiently computing the eigenvalues of limited-memory quasi-Newton matrices
• On solving L-SR1 trust-region subproblems
• On the Behavior of Broyden’s Class of Quasi-Newton Methods
• On the Inverse of the Sum of Matrices
• Quasi-Newton Algorithms with Updates from the Preconvex Part of Broyden's Family
• Quasi-Newton Methods, Motivation and Theory
• Representations of quasi-Newton matrices and their use in limited memory methods
• Statistical Quasi-Newton: A New Look at Least Change
• The linear algebra of block quasi-Newton algorithms
• Trust Region Methods
• Updating Quasi-Newton Matrices with Limited Storage

Cited in

(10)

• Automatic Preconditioning by Limited Memory Quasi-Newton Updating
• Quantitative performance analysis of the improved quasi-minimal residual method on massively distributed memory computers.
• Limited-memory BFGS systems with diagonal updates
• Representations of quasi-Newton matrices and their use in limited memory methods
• Exact linesearch limited-memory quasi-Newton methods for minimizing a quadratic function
• scientific article; zbMATH DE number 940915 (Why is no real title available?)
• On efficiently computing the eigenvalues of limited-memory quasi-Newton matrices
• An autoadaptative limited memory Broyden's method to solve systems of nonlinear equations
• A dense initialization for limited-memory quasi-Newton methods
• Shifted L-BFGS systems

Describes a project that uses

Uses Software

• L-BFGS

This page was built for publication: On solving large-scale limited-memory quasi-Newton equations

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q501277)