Preprocessing Data for Machine Learning Models