Causal Inference and Machine Learning for Economic Applications
AAEC 8810
3 hours
Causal Inference and Machine Learning for Economic Applications
Analytical Thinking
Critical Thinking
Course Description
The theory and application of causal machine learning. It will be divided into three components: the first component will review the basics of machine learning (ML) and commonly used ML algorithms; the second will introduce the causal inference framework; and finally, in the third part, we will combine the two for data-driven heterogeneous treatment effects and policy estimation and evaluation. The course will also involve data applications using R for each component.
Athena Title
Causal ML for Economics
Prerequisite
ECON 8080
Semester Course Offered
Offered fall
Grading System
A - F (Traditional)
Student learning Outcomes
By the end of this course, students will learn about the commonly used Machine Learning (ML) methods.
By the end of this course, students will apply ML methods to economic data to estimate models that predict economic outcomes well.
By the end of this course, students will learn the theory and application of policy learning that uses both causal inference and ML.
By the end of this course, students will estimate and evaluate individualized treatment assignment rules on both randomized data and observational studies using R.
Topical Outline
1. Introduction to Machine Learning
a. Basics such as modeling, fitting and estimation
b. ML methods- supervised and unsupervised learning penalized regressions, decision trees, forests, clustering, matrix completion algorithms
2. Average Treatment Effects
a. Potential Outcomes framework
b. Estimands
c. Estimators- Difference in Means Estimator, Inverse Propensity Weighted Estimator, Augmented Inverse Propensity Weighted Estimator
3. Heterogeneity in Treatment Effects
a. CATE (Conditional Average Treatment Effect)
b. Heterogeneity for pre-specified hypothesis
c. Heterogeneity for data driven hypothesis
4. Policy Learning and Evaluation
a. Basics- estimands, estimators, cross fitting
b. Parametric and non parametric policies
c. Policy estimation on randomized data and on observational studies
d. Estimation via LASSO and causal forest
e. Policy evaluation-off policy and on policy
f. Policy without and with budget constraints
Institutional Competencies Learning Outcomes
Analytical Thinking
The ability to reason, interpret, analyze, and solve problems from a wide array of authentic contexts.
Critical Thinking
The ability to pursue and comprehensively evaluate information before accepting or establishing a conclusion, decision, or action.
