Please use this identifier to cite or link to this item:
Title: Using sampling and simplex derivatives in pattern search methods
Authors: Custódio, Ana Luísa 
Vicente, Luís Nunes 
Keywords: Derivative free optimization; Pattern search methods; Simplex gradient; Simplex Hessian; Multivariate polynomial interpolation; Poisedness
Issue Date: 2004
Publisher: Centro de Matemática da Universidade de Coimbra
Citation: Pré-Publicações DMUC. 04-35 (2004)
Abstract: Pattern search methods can be made more efficient if past function evaluations are appropriately reused. In this paper we will introduce a number of ways of reusing previous evaluations of the objective function based on the computation of simplex derivatives (e.g., simplex gradients) to improve the efficiency of a pattern search iteration. At each iteration of a pattern search method, one can attempt to compute an accurate simplex gradient by identifying a sampling set of previous iterates with good geometrical properties. This simplex gradient computation can be done using only past successful iterates or by considering all past function evaluations. The simplex gradient can then be used, for instance, to reorder the evaluations of the objective function associated with the positive spanning set or positive basis used in the poll step. But it can also be used to update the mesh size parameter according to a sufficient decrease criterion. None of these modifications demands new function evaluations. A search step can also be tried along the negative simplex gradient at the beginning of the current pattern search iteration. We will present these procedures in detail and show how promising they are to enhance the practical performance of pattern search methods.
Rights: openAccess
Appears in Collections:FCTUC Matemática - Artigos em Revistas Nacionais

Files in This Item:
File Description SizeFormat
Using sampling and simplex derivatives.pdf237.88 kBAdobe PDFView/Open
Show full item record

Page view(s) 20

checked on Sep 22, 2020


checked on Sep 22, 2020

Google ScholarTM


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.