中文課名：經濟學與政治科學中的因果機器學習
授課教師：陳釗而
研究室：台大次震宇宙館 223室
Course Description
This course introduces machine learning methods and their applications to causal inference in economics and political science. In data-rich empirical research, the central challenge lies not only in asking meaningful questions but also in applying empirical methods with rigor and good judgment. To this end, we examine examples of off-the-shelf machine learning tools used in economics and political science, followed by highlights from the emerging econometric literature that integrates machine learning with causal inference.
Students are expected to complete all problem sets and assigned readings. Active participation is strongly encouraged—questions and class discussion will be an integral part of the learning process. Mastery of the techniques taught in this course will be evaluated through four assignments and one in-class final exam.
Course Objective
By the end of the course, students will have developed a working familiarity with causal machine learning techniques as well as practical skills in data handling and programming.
Course Outline
- Review of statistics; identification; the potential outcomes framework
- The Furious Five – Randomized controlled trials
- The Furious Five – Regression, Part I
- The Furious Five – Regression (“Matchmaker”), Part II
- The Furious Five – Instrumental variables
- The Furious Five – Difference-in-differences
- The Furious Five – Regression discontinuity design
- Panel data models and synthetic control methods
- A helicopter tour of causal machine learning in economics and political science
- Modern high-dimensional econometrics
- The double-lasso selection procedure
- Decision trees and random forests
- Causal forests
- Heterogeneous treatment effects and policy learning
- Review and discussion
- Review and discussion
Readings
- Angrist, J., and Pischke, J. (2009). Mostly Harmless Econometrics. Princeton University Press.
- James, G., Witten, D., Hastie, T., and Tibshirani, R. (2021). An Introduction to Statistical Learning with Applications in R, 2nd ed. Springer.
- Chen, Jau-er, and Jing, Annette (2025). “Recent Advances in Causal Machine Learning and Dynamic Policy Learning.” Wiley Interdisciplinary Reviews: Computational Statistics, 17(4), 1–27.
Grading
Assessment consists of four assignments (80%) and one in-class final exam (20%).
Jauer 