The course objective is to introduce the basic concepts in predictive analytics, also called predictive modeling, the most prevalent form of data mining. This course covers the two core paradigms that account for most business applications of predictive modeling: classification and prediction. Students learn how to explore and visualize data to get a preliminary idea of what variables are important, and how they relate to one another. Four machine learning techniques are used: k-nearest neighbors, classification, and regression trees (CART), and Bayesian classifiers. Students then learn how to combine different models to obtain results that are better than any of the individual models produce on their own. The course also covers the use of partitioning to divide the data into training data (data used to build a model), validation data (data used to assess the performance of different models, or, in some cases, to fine tune the model) and test data (data used to predict the performance of the final model).