This course explores what are the various issues that arise when machine and statistical learning methods are used in practice on big data to inform business intelligence (in finance and insurance). In practice, data is not clean, number of features is large, feature engineering must be carried out, and data is often multi-modal consisting not only of structured data, but also of images, text, and social network data. In this course, students will be exposed to various techniques and practical know-how to deal with these cases and learn how to present results to practitioners who are not domain experts.