Optimizing partially separable functions without derivatives

Résultats de recherche: Contribution à un journal/une revueArticle

49 Downloads (Pure)

Résumé

We present an algorithm for solving nonlinear programming problems involving a partially separable objective function whose derivatives are assumed to be unavailable. At each iteration, we construct a quadratic interpolation model of the objective function around the current iterate and minimize this model to obtain a trial step. The whole process is embedded within a trust-region framework. We further propose to use ideas of Curtis, Powell and Reid to minimize the number of calls to the objective function in the part of the derivative-free algorithm that improves the geometry of the interpolation set. Numerical experiments tend to confirm the promising behaviour of the algorithm. © 2005 Taylor & Francis Group Ltd.
langue originaleAnglais
Pages (de - à)493-508
Nombre de pages16
journalOptimization Methods and Software
Volume20
Numéro de publication4-5
Les DOIs
étatPublié - 1 août 2005

Empreinte digitale

Objective function
Derivatives
Derivative
Interpolation
Interpolate
Minimise
Derivative-free
Trust Region
Nonlinear programming
Set theory
Nonlinear Programming
Iterate
Numerical Experiment
Tend
Iteration
Geometry
Model
Experiments
Framework

Citer ceci

@article{390805fbb93c4e149905559d4e9360d7,
title = "Optimizing partially separable functions without derivatives",
abstract = "We present an algorithm for solving nonlinear programming problems involving a partially separable objective function whose derivatives are assumed to be unavailable. At each iteration, we construct a quadratic interpolation model of the objective function around the current iterate and minimize this model to obtain a trial step. The whole process is embedded within a trust-region framework. We further propose to use ideas of Curtis, Powell and Reid to minimize the number of calls to the objective function in the part of the derivative-free algorithm that improves the geometry of the interpolation set. Numerical experiments tend to confirm the promising behaviour of the algorithm. {\circledC} 2005 Taylor & Francis Group Ltd.",
author = "B. Colson and Philippe Toint",
note = "Copyright 2008 Elsevier B.V., All rights reserved.",
year = "2005",
month = "8",
day = "1",
doi = "10.1080/10556780500140227",
language = "English",
volume = "20",
pages = "493--508",
journal = "Optimization Methods and Software",
issn = "1055-6788",
publisher = "Taylor & Francis",
number = "4-5",

}

Optimizing partially separable functions without derivatives. / Colson, B.; Toint, Philippe.

Dans: Optimization Methods and Software, Vol 20, Numéro 4-5, 01.08.2005, p. 493-508.

Résultats de recherche: Contribution à un journal/une revueArticle

TY - JOUR

T1 - Optimizing partially separable functions without derivatives

AU - Colson, B.

AU - Toint, Philippe

N1 - Copyright 2008 Elsevier B.V., All rights reserved.

PY - 2005/8/1

Y1 - 2005/8/1

N2 - We present an algorithm for solving nonlinear programming problems involving a partially separable objective function whose derivatives are assumed to be unavailable. At each iteration, we construct a quadratic interpolation model of the objective function around the current iterate and minimize this model to obtain a trial step. The whole process is embedded within a trust-region framework. We further propose to use ideas of Curtis, Powell and Reid to minimize the number of calls to the objective function in the part of the derivative-free algorithm that improves the geometry of the interpolation set. Numerical experiments tend to confirm the promising behaviour of the algorithm. © 2005 Taylor & Francis Group Ltd.

AB - We present an algorithm for solving nonlinear programming problems involving a partially separable objective function whose derivatives are assumed to be unavailable. At each iteration, we construct a quadratic interpolation model of the objective function around the current iterate and minimize this model to obtain a trial step. The whole process is embedded within a trust-region framework. We further propose to use ideas of Curtis, Powell and Reid to minimize the number of calls to the objective function in the part of the derivative-free algorithm that improves the geometry of the interpolation set. Numerical experiments tend to confirm the promising behaviour of the algorithm. © 2005 Taylor & Francis Group Ltd.

UR - http://www.scopus.com/inward/record.url?scp=27744469963&partnerID=8YFLogxK

U2 - 10.1080/10556780500140227

DO - 10.1080/10556780500140227

M3 - Article

VL - 20

SP - 493

EP - 508

JO - Optimization Methods and Software

JF - Optimization Methods and Software

SN - 1055-6788

IS - 4-5

ER -