Abstract
The problem of setting the parameter values of a metaheuristic algorithm that optimise its performance is complex and time-consuming. Although the performance of a metaheuristic can be very sensitive to the parameter values, it is usual in the literature that the selection of the value parameters is not enough justified. There are in the literature two procedures that facilitate the task of fine-tuning: CALIBRA and the Nelder & Mead (N&M) algorithm. We propose a hands-off systematic procedure for fine-tuning metaheuristics that takes the advantages of CALIBRA and the N&M algorithm.
Acknowledgements
The authors gratefully acknowledge the support of grant DPI2007-61905 (Ministry of Education and Science, Spain, and FEDER). The authors also express their gratitude to the anonymous reviewers for their accurate reviews and valuable comments, which have helped to improve the quality of this paper.