A nonlinear preconditioner for optimum experimental design problems

We show how to efficiently compute A-optimal experimental designs, which are formulated in terms of the minimization of the trace of the covariance matrix of the underlying regression process, using quasi-Newton sequential quadratic programming methods. In particular, we introduce a modification of...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Mario S. Mommer, Andreas Sommer, Johannes P. Schlöder, H. Georg Bock
Formato: article
Lenguaje:EN
Publicado: Elsevier 2015
Materias:
Acceso en línea:https://doaj.org/article/5227be4f681c41179ccb2281a60abe78
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
id oai:doaj.org-article:5227be4f681c41179ccb2281a60abe78
record_format dspace
spelling oai:doaj.org-article:5227be4f681c41179ccb2281a60abe782021-12-02T05:00:43ZA nonlinear preconditioner for optimum experimental design problems2192-440610.1007/s13675-015-0036-9https://doaj.org/article/5227be4f681c41179ccb2281a60abe782015-05-01T00:00:00Zhttp://www.sciencedirect.com/science/article/pii/S2192440621000423https://doaj.org/toc/2192-4406We show how to efficiently compute A-optimal experimental designs, which are formulated in terms of the minimization of the trace of the covariance matrix of the underlying regression process, using quasi-Newton sequential quadratic programming methods. In particular, we introduce a modification of the problem that leads to significantly faster convergence. To derive this modification, we model each iteration in terms of an initial experimental design that is to be improved, and show that the absolute condition number of the model problem grows without bounds as the quality of the initial design improves. As a remedy, we devise a preconditioner that ensures that the absolute condition number will instead stay uniformly bounded. Using numerical experiments, we study the effect of this reformulation on relevant cases of the general problem class, and find that it leads to significant improvements in both stability and convergence behavior.Mario S. MommerAndreas SommerJohannes P. SchlöderH. Georg BockElsevierarticle90C3090C5562K99Applied mathematics. Quantitative methodsT57-57.97Electronic computers. Computer scienceQA75.5-76.95ENEURO Journal on Computational Optimization, Vol 3, Iss 2, Pp 131-146 (2015)
institution DOAJ
collection DOAJ
language EN
topic 90C30
90C55
62K99
Applied mathematics. Quantitative methods
T57-57.97
Electronic computers. Computer science
QA75.5-76.95
spellingShingle 90C30
90C55
62K99
Applied mathematics. Quantitative methods
T57-57.97
Electronic computers. Computer science
QA75.5-76.95
Mario S. Mommer
Andreas Sommer
Johannes P. Schlöder
H. Georg Bock
A nonlinear preconditioner for optimum experimental design problems
description We show how to efficiently compute A-optimal experimental designs, which are formulated in terms of the minimization of the trace of the covariance matrix of the underlying regression process, using quasi-Newton sequential quadratic programming methods. In particular, we introduce a modification of the problem that leads to significantly faster convergence. To derive this modification, we model each iteration in terms of an initial experimental design that is to be improved, and show that the absolute condition number of the model problem grows without bounds as the quality of the initial design improves. As a remedy, we devise a preconditioner that ensures that the absolute condition number will instead stay uniformly bounded. Using numerical experiments, we study the effect of this reformulation on relevant cases of the general problem class, and find that it leads to significant improvements in both stability and convergence behavior.
format article
author Mario S. Mommer
Andreas Sommer
Johannes P. Schlöder
H. Georg Bock
author_facet Mario S. Mommer
Andreas Sommer
Johannes P. Schlöder
H. Georg Bock
author_sort Mario S. Mommer
title A nonlinear preconditioner for optimum experimental design problems
title_short A nonlinear preconditioner for optimum experimental design problems
title_full A nonlinear preconditioner for optimum experimental design problems
title_fullStr A nonlinear preconditioner for optimum experimental design problems
title_full_unstemmed A nonlinear preconditioner for optimum experimental design problems
title_sort nonlinear preconditioner for optimum experimental design problems
publisher Elsevier
publishDate 2015
url https://doaj.org/article/5227be4f681c41179ccb2281a60abe78
work_keys_str_mv AT mariosmommer anonlinearpreconditionerforoptimumexperimentaldesignproblems
AT andreassommer anonlinearpreconditionerforoptimumexperimentaldesignproblems
AT johannespschloder anonlinearpreconditionerforoptimumexperimentaldesignproblems
AT hgeorgbock anonlinearpreconditionerforoptimumexperimentaldesignproblems
AT mariosmommer nonlinearpreconditionerforoptimumexperimentaldesignproblems
AT andreassommer nonlinearpreconditionerforoptimumexperimentaldesignproblems
AT johannespschloder nonlinearpreconditionerforoptimumexperimentaldesignproblems
AT hgeorgbock nonlinearpreconditionerforoptimumexperimentaldesignproblems
_version_ 1718400827785216000