Recent progress in unconstrained nonlinear optimization without derivatives

A.R. Conn, K. Scheinberg, Ph.L. Toint

Résultats de recherche: Contribution à un journal/une revueArticle

42 Downloads (Pure)

Résumé

We present an introduction to a new class of derivative free methods for unconstrained optimization. We start by discussing the motivation for such methods and why they are in high demand by practitioners. We then review the past developments in this field, before introducing the features that characterize the newer algorithms. In the context of a trust region framework, we focus on techniques that ensure a suitable "geometric quality" of the considered models. We then outline the class of algorithms based on these techniques, as well as their respective merits. We finally conclude the paper with a discussion of open questions and perspectives. © 1997 The Mathematical Programming Society, Inc. Published by Elsevier Science B.V.
langue originaleAnglais
Pages (de - à)397-414
Nombre de pages18
journalMathematical Programming Series B
Volume79
Numéro de publication3
étatPublié - 1 oct. 1997

Empreinte digitale

Unconstrained Optimization
Nonlinear Optimization
Derivative-free Methods
Derivatives
Derivative
Trust Region
Mathematical programming
Mathematical Programming
Class
Model
Demand
Context
Review
Framework

Citer ceci

@article{297952882b2f49319a5a9e6871830449,
title = "Recent progress in unconstrained nonlinear optimization without derivatives",
abstract = "We present an introduction to a new class of derivative free methods for unconstrained optimization. We start by discussing the motivation for such methods and why they are in high demand by practitioners. We then review the past developments in this field, before introducing the features that characterize the newer algorithms. In the context of a trust region framework, we focus on techniques that ensure a suitable {"}geometric quality{"} of the considered models. We then outline the class of algorithms based on these techniques, as well as their respective merits. We finally conclude the paper with a discussion of open questions and perspectives. {\circledC} 1997 The Mathematical Programming Society, Inc. Published by Elsevier Science B.V.",
author = "A.R. Conn and K. Scheinberg and Ph.L. Toint",
note = "Publication code : {"}*FP SB 10/1997/12 ; QA 2.2/1/97/12{"}",
year = "1997",
month = "10",
day = "1",
language = "English",
volume = "79",
pages = "397--414",
journal = "Mathematical Programming",
issn = "0025-5610",
publisher = "Springer-Verlag GmbH and Co. KG",
number = "3",

}

Recent progress in unconstrained nonlinear optimization without derivatives. / Conn, A.R.; Scheinberg, K.; Toint, Ph.L.

Dans: Mathematical Programming Series B, Vol 79, Numéro 3, 01.10.1997, p. 397-414.

Résultats de recherche: Contribution à un journal/une revueArticle

TY - JOUR

T1 - Recent progress in unconstrained nonlinear optimization without derivatives

AU - Conn, A.R.

AU - Scheinberg, K.

AU - Toint, Ph.L.

N1 - Publication code : "*FP SB 10/1997/12 ; QA 2.2/1/97/12"

PY - 1997/10/1

Y1 - 1997/10/1

N2 - We present an introduction to a new class of derivative free methods for unconstrained optimization. We start by discussing the motivation for such methods and why they are in high demand by practitioners. We then review the past developments in this field, before introducing the features that characterize the newer algorithms. In the context of a trust region framework, we focus on techniques that ensure a suitable "geometric quality" of the considered models. We then outline the class of algorithms based on these techniques, as well as their respective merits. We finally conclude the paper with a discussion of open questions and perspectives. © 1997 The Mathematical Programming Society, Inc. Published by Elsevier Science B.V.

AB - We present an introduction to a new class of derivative free methods for unconstrained optimization. We start by discussing the motivation for such methods and why they are in high demand by practitioners. We then review the past developments in this field, before introducing the features that characterize the newer algorithms. In the context of a trust region framework, we focus on techniques that ensure a suitable "geometric quality" of the considered models. We then outline the class of algorithms based on these techniques, as well as their respective merits. We finally conclude the paper with a discussion of open questions and perspectives. © 1997 The Mathematical Programming Society, Inc. Published by Elsevier Science B.V.

UR - http://www.scopus.com/inward/record.url?scp=0242609515&partnerID=8YFLogxK

M3 - Article

VL - 79

SP - 397

EP - 414

JO - Mathematical Programming

JF - Mathematical Programming

SN - 0025-5610

IS - 3

ER -