Data Science - Statistical Learning I

Catalog description

Statistical learning is a fundamental skill for data scientists. Data scientists are specialists in "drinking from the firehose" of big data, and statistical learning techniques are some of their key tools. This course focuses on applications of statistical learning to big data challenges through data mining and predictive modeling techniques that are in great demand. Students will be introduced to the basics of statistical/machine learning: supervised learning (e.g. linear model, nonlinear models, penalized methods, ensemble methods, etc.), unsupervised learning (e.g. K means clustering, nearest neighbors, hierarchical clustering, etc.), and missing data in machine learning. Throughout the course, we will learn how to be "informed doers", who not only know how to apply methods but understand how those methods work. This understanding can be critical to getting good results from big data, so that the limitations of certain methods are properly understood. Prerequisite: STAT 820 or STAT 823, STAT 835, STAT 840, or by permission of instructor.