A class of diagonal preconditioners for limited memory BFGS method.

A major weakness of the limited memory BFGS (LBFGS) method is that it may converge very slowly on ill-conditioned problems when the identity matrix is used for initialization. Very often, the LBFGS method can adopt a preconditioner on the identity matrix to speed up the convergence. For this purpo...

Full description

Bibliographic Details
Main Authors: Leong, Wah June, Chen, Chuei Yee
Format: Article
Language:English
English
Published: Taylor & Francis 2013
Online Access:http://psasir.upm.edu.my/id/eprint/29993/
http://psasir.upm.edu.my/id/eprint/29993/1/A%20class%20of%20diagonal%20preconditioners%20for%20limited%20memory%20BFGS%20method.pdf
_version_ 1848846552917868544
author Leong, Wah June
Chen, Chuei Yee
author_facet Leong, Wah June
Chen, Chuei Yee
author_sort Leong, Wah June
building UPM Institutional Repository
collection Online Access
description A major weakness of the limited memory BFGS (LBFGS) method is that it may converge very slowly on ill-conditioned problems when the identity matrix is used for initialization. Very often, the LBFGS method can adopt a preconditioner on the identity matrix to speed up the convergence. For this purpose, we propose a class of diagonal preconditioners to boost the performance of the LBFGS method. In this context, we find that it is appropriate to use a diagonal preconditioner, in the form of a diagonal matrix plus a positive multiple of the identity matrix, so as to fit information of local Hessian as well as to induce positive definiteness for the diagonal preconditioner at a whole. The property of hereditary positive definiteness is maintained by a careful choice of the positive scalar on the scaled identity matrix while the local curvature information is carried implicitly on the other diagonal matrix through the variational techniques, commonly employed in the derivation of quasi-Newton updates. Several preconditioning formulae are then derived and tested on a large set of standard test problems to access the impact of different choices of such preconditioners on the minimization performance.
first_indexed 2025-11-15T09:04:32Z
format Article
id upm-29993
institution Universiti Putra Malaysia
institution_category Local University
language English
English
last_indexed 2025-11-15T09:04:32Z
publishDate 2013
publisher Taylor & Francis
recordtype eprints
repository_type Digital Repository
spelling upm-299932015-10-01T23:38:32Z http://psasir.upm.edu.my/id/eprint/29993/ A class of diagonal preconditioners for limited memory BFGS method. Leong, Wah June Chen, Chuei Yee A major weakness of the limited memory BFGS (LBFGS) method is that it may converge very slowly on ill-conditioned problems when the identity matrix is used for initialization. Very often, the LBFGS method can adopt a preconditioner on the identity matrix to speed up the convergence. For this purpose, we propose a class of diagonal preconditioners to boost the performance of the LBFGS method. In this context, we find that it is appropriate to use a diagonal preconditioner, in the form of a diagonal matrix plus a positive multiple of the identity matrix, so as to fit information of local Hessian as well as to induce positive definiteness for the diagonal preconditioner at a whole. The property of hereditary positive definiteness is maintained by a careful choice of the positive scalar on the scaled identity matrix while the local curvature information is carried implicitly on the other diagonal matrix through the variational techniques, commonly employed in the derivation of quasi-Newton updates. Several preconditioning formulae are then derived and tested on a large set of standard test problems to access the impact of different choices of such preconditioners on the minimization performance. Taylor & Francis 2013 Article PeerReviewed application/pdf en http://psasir.upm.edu.my/id/eprint/29993/1/A%20class%20of%20diagonal%20preconditioners%20for%20limited%20memory%20BFGS%20method.pdf Leong, Wah June and Chen, Chuei Yee (2013) A class of diagonal preconditioners for limited memory BFGS method. Optimization Methods and Software, 28 (2). pp. 379-392. ISSN 1055-6788; ESSN: 1029-4937 10.1080/10556788.2011.653356 English
spellingShingle Leong, Wah June
Chen, Chuei Yee
A class of diagonal preconditioners for limited memory BFGS method.
title A class of diagonal preconditioners for limited memory BFGS method.
title_full A class of diagonal preconditioners for limited memory BFGS method.
title_fullStr A class of diagonal preconditioners for limited memory BFGS method.
title_full_unstemmed A class of diagonal preconditioners for limited memory BFGS method.
title_short A class of diagonal preconditioners for limited memory BFGS method.
title_sort class of diagonal preconditioners for limited memory bfgs method.
url http://psasir.upm.edu.my/id/eprint/29993/
http://psasir.upm.edu.my/id/eprint/29993/
http://psasir.upm.edu.my/id/eprint/29993/1/A%20class%20of%20diagonal%20preconditioners%20for%20limited%20memory%20BFGS%20method.pdf