Choice and Forecasting
Models of Choice and Forecasting is intended to augment a first course in linear regression with extensions of linear models to generalized linear models for a broader array of data types and a dive into time series regression and forecasting.
The first half of the course is built around The Handbook of Regression Modeling in People Analytics: With Examples in R, Python, and Julia by Keith McNulty [Global Director of Talent Sciences at McKinsey and Company] that covers models for discrete data types [binary, ordered, nominal choices, counts of events, and survival/duration analysis].
The second half of the course relies on the excellent Forecasting, Principles and Practice, 3rd Edition, by Rob J. Hyndman and George Athanasopoulos of Monash University in Australia that is entirely supported by R libraries for time series problems.
The lectures will focus on intuition and the mathematical logic but the goal is to put the tools into practice. To this end, the expectations are weekly homework exercises to insure that we can actually do what we are presented but there are two key summary deliverables: a project employing a detailed application of choice models near the middle and a project in time series forecasting due at the end of the term. Both are to be presented at the end of the term.