Practical initialization of the Nelder–Mead method for computationally expensive optimization problems

Abstract Black-box optimization (BBO) algorithms are widely employed by practitioners to address computationally expensive real-world problems such as automatic tuning of machine learning models and evacuation route planning. The Nelder–Mead (NM) method is a well-known local search heuristic for BBO that has been applied solve many from way back because its promising performance. However, this strong dependence on initialization due tendency. Nevertheless, discussion the proper NM limited recent study Wessing (Optim Lett 13(4):847–856, 2019), which solely based an analysis using simple sphere function. In study, we take further step improve Wessing’s result massively investigating how affects performance in views initial simplex size shape constraint handling 24 benchmarking problems. Based numerical results, present empirical best practice cases involving evaluation budget.