Description: Lecture, three hours; discussion, one hour. Limited to Master of Applied Statistics students. Often we are interested in making inferences and predictions from data, either by (1) estimating particular meaningful parameters of models or (2) finding best fitting model that we can then manipulate to produce useful outputs such as predictions or counterfactual estimates. Focus on what is done when linear models are not appropriate and may produce misleading estimates. Generalized linear model and maximum likelihood methods as essential tools all statistics students should understand. Examination of shift gears to explore regression and classification techniques that have been ubiquitous in machine learning literature in recent years, with special attention to regularization and kernelized methods. S/U or letter grading.

Units: 4.0