Skip to the main content

Professional paper

The Nelder-Mead method: a local method for direct unconstrained optimization

Lucijana Grgić ; Odjel za matematiku, Sveučilište J.J. Strossmayera u Osijeku, Osijek
Kristian Sabo ; Odjel za matematiku, Sveučilište J.J. Strossmayera u Osijeku, Osijek


Full text: croatian pdf 1.868 Kb

page 131-143

downloads: 662

cite


Abstract

In this paper, we describe the popular Nelder–Mead method, which
is considered to be one of the best known local methods of direct unconstrained
optimization. For the sake of simplicity, we analize a special
case of optimization in \(\mathbb{R}^2\)
since in that case the Nelder–Mead method
boiles down to a sequence of elementary geometric transformations
in the plane such that, for a complete understanding thereof, highschool
mathematics suffices. In order to illustrate the method, we give
several numerical examples using Mathematica

Keywords

unconstrained optimization; local optimization; direct method; Nelder–Mead method

Hrčak ID:

161398

URI

https://hrcak.srce.hr/161398

Publication date:

31.3.2016.

Article data in other languages: croatian

Visits: 1.831 *