After that, the data is cleaned and ready for analysis through preprocessing. This could be working with migame cash appssing values, eliminating outliers, or formatting the data so that it can be analyzed properly. After preprocessing the data, the predictive app trains a model on historical data using machine learning algorithms.