Regression for Y-hat

slides

This class is about regression as a tool to approximate a conditional expectation function. From this perspective, the \(\hat\beta\) estimates are only a step toward the broader purpose of regression to produce \(\hat{Y}\) values that achieve this approximation well. This perspective will ultimately allow us to consider machine learning estimators beyond regression.

Some of this class relies on an ongoing project on description: ilundberg.github.io/description