Comparisons of Machine Learning Classification Models with Tidymodels

This chapter focuses on building and comparing six machine learning classification models: (1) logistic regression, (2) Least Absolute Shrinkage and Selection Operator (LASSO) logistic regression, (3) support vector machine, (4) random forest, (5) extreme gradient boosted trees, and (6) K-nearest neighbor using the tidymodels framework. In addition to the Summary, there are three sections. Section 19.1 illustrates how to prepare data for machine learning, including (1) importing and preparing the dataset, (2) splitting the data into training and testing datasets, and (3) creating a validation dataset. Section 19.2 introduces machine learning step by step with tidymodels, including (1) data preprocessing recipes, (2) building models, (3) creating workflows, (4) evaluating models, (5) comparing models, (6) fitting the last evaluation on the test dataset, and (7) illustrating details on model fitting. Section 19.3 provides remarks on comparisons of machine learning models using F 1 score, Matthews correlation coefficient (MCC), area under the receiver operating characteristic curve (AUC-ROC) and area under the precision-recall curve (AUC-PR), and particularly comments on F 1 score, MCC, and AUC-ROC metrics.