Global optimization approaches for parameter tuning in biomedical signal processing: A focus on multi-scale entropy

Many algorithms used for the analysis of physiological signals include hyper-parameters that must be selected by the investigator. The ultimate choice of these parameter values can have a dramatic impact on the performance of the approach. Addressing this issue often requires investigators to manually tune parameters for their particular data-set. In this study, we illustrate the importance of global optimization techniques for the automated determination of parameter values in the multi-scale entropy (MSE) algorithm. Importantly, we demonstrate that global optimization techniques provide an effective, and automated framework for tuning parameters of such algorithms, and easily improve upon the default settings selected by experts.