Optimizing the parameters of a physical exercise dose-response model

Abstract

The purpose of this research was to compare the robustness and performance of a local and global optimization algorithm applied to the problem of fitting the parameters of a non-linear dose-response model utilized in the field of exercise physiology. Traditionally the parameters of dose-response models utilised in exercise physiology have been fit with non-linear least squares procedures in combination with local optimization algorithms. These algorithms have demonstrated limitations in their ability to converge on a globally optimal solution. This research purposes the use of an evolutionary computation based algorithm as an alternative method to fit a nonlinear dose-response model. The results of our comparison over 1000 experimental runs demonstrated the superior performance of the evolutionary computation based algorithm to consistently achieve a more consistent model fit and holdout evaluation performance in comparison to the local search algorithm. This initial research would suggest that global evolutionary computation based optimization algorithms are a fast and more robust alternative to local optimization algorithms when fitting the parameters of nonlinear dose-response models.