Funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Education and Culture Executive Agency (EACEA). Neither the European Union nor EACEA can be held responsible for them.

Library

Derivative-free optimization: a review of algorithms and comparison of software implementation

Users: 1 - Average Rating: 5.00


This work provides an extensive overview of techniques and outlines well-established methods with respect to derivative-free optimisation algorithms, with pointers to more detailed surveys and an extensive catalogue of key developments: The emergence of derivative-free optimization, the appearance of simulation test-beds for comparing algorithms, the application of these methods in various domains of science and engineering.
In addition, this work provides a comprehensive benchmark of 22 different and well-known solvers in various constrained and unconstrained, convex and non-convex problems, ranging between 1 – 300 decision variables. The authors demonstrate that attaining the best solutions even for small problems still remains a challenge for most current derivative-free solvers. In addition, they conclude that there is no single solver whose performance dominates that of all others, and that all solvers provided the best solution possible for at least some of the test problems.

Type:
Scientific Paper

Area:
Optimization

Target Group:
Advanced

DOI:
10.1007/s10898-012-9951-y


Cite as:
Rios, Luis Miguel, and Nikolaos V. Sahinidis. "Derivative-free optimization: a review of algorithms and comparison of software implementations." Journal of Global Optimization 56.3 (2013): 1247-1293.

Author of the review:
Miltiadis Kalikatzarakis
University of Strathclyde


Reviews

You have to login to leave a comment. If you are not registered click here

Eligius Hendrix


The mentioned paper excells in doing a very systematic investigation on derivative free methods. In the end you want to know which method is more fit for which type of optimization problem.