Teaching

Machine Learning for Policy Analysis

SPIA 586a · Spring 2024, 2025, 2027

Machine Learning for Policy Analysis is a graduate course on using machine learning methods in policy-relevant empirical work, with attention to discovery, measurement, prediction, designed and found data, privacy, fairness, interpretability, and translating models into practice.

Materials

2025 Lecture Slides

January 27/29

Week 1: Introduction

February 3/5

Week 2: Discovery and Measurement

February 10/12

Week 3: Prediction Within Population

February 17/19

Week 4: Prediction in a New Population

February 24/26

Week 5: Prediction to a Counterfactual Population

March 3/5

Week 6: Experimental Data

March 10/12

Spring Break

March 17/19

Week 7: Designed Data

March 24/26

Week 8: Found Data

March 31/April 2

Week 9: Privacy

April 7/9

Week 10: Fairness

April 14/16

Week 11: Interpretability

April 21/23

Week 12: Machine Learning in Practice and Policy

2024

Syllabus