Graduate students Maryam Feyzollahi and Nima Rafizadeh have teamed up on several research projects, including a new guide that helps economists bring machine learning into causal analysis with confidence.
"Double/Debiased Machine Learning for Economists: Practical Guidelines, Best Practices, and Common Pitfalls"
Double/debiased machine learning lets economists use machine learning to answer causal questions, but clear, practical guidance has been scarce. Maryam and Nima fill that gap with best practices for choosing and evaluating models, and they clear up a common confusion between cross-validation, which is for choosing a model, and cross-fitting, which is for getting unbiased estimates.