For machine learning methods, Zhang et al used Bayesian optimization to select the optimal combination of hyperparameters for different models (e.g., support vector regression, decision tree, random forest, etc.) for penetration rate prediction. Wang et al