source: The Economist
Instructor:
6 Regularization
Readings
Intro to Statistical Learning 6.1, 6.2
Using Post-Double Selection Lasso in Field Experiments by Jacobus Cilliers, Nour Ahmed Mohamed Abdelwahab Elashmawy, and David Mckenzie
Lecture
Lab
The objective is to use lasso to select covariates for the EMERGE project. We will use a data set containing 200 observations of 25 variables representing household characteristics and measures of child development. Additional information on the study is available here.
For this lab, the R and Python implementions are not identical; Python requires the user to handle a number of steps that are done under the hood in R. I therefore recommend using either the R version or the Python version of the lab template rather than than the generic version.