Skip to content

MATLAB implementation of a few variants of the evolution strategy optimisation technique which are then tested on the 5D Rana function

Notifications You must be signed in to change notification settings

CharlesArnal/Evolution_strategies_and_tabu_search_optimisation

Repository files navigation

Black box optimisation using evolution strategies and tabu search

MATLAB implementation of a few variants of the evolution strategy and tabu search optimisation techniques which are tested on the 5D Rana function.

Detailed information, theoretical discussion and experimental results can be found in "ES_TS_report.pdf".

This was done in the context of the University of Cambridge's Practical Optimisation course (Engineering Tripos, Part IIB), under the supervision of Dr. Geoff Parks.

Evolution Strategy

We test various recombination and mutation operators (including some personal inventions) and choices of parameters, as well as the Covariance matrix adaptation ES (CMA-ES) technique. Example of experiments can be found in "Experiment_ES.m" and "Experiment_CMA_ES.m".

Tabu Search

We test various choices of parameters, as well as two variations on the idea of Concentric Tabu Search. Example of experiments can be found in "Experiment_Tabu.m".

About

MATLAB implementation of a few variants of the evolution strategy optimisation technique which are then tested on the 5D Rana function

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages