This study aims to compare the performance of linear and nonlinear models in data-poor and data-rich environments with different regularization and loss functions to understand which characteristics of machine learning are ...