Machine Learning for Economics

The objective of this course is to study machine learning methods and their practical applications in empirical economic research and public policy design. The methods covered include regression and classification trees, bagging, random forests, boosting, neural networks (deep learning), and bandit algorithms. We also examine their applications to statistical causal inference, public policy design, and macroeconomic forecasting. In addition, we discuss the use of alternative data, such as images and text, in empirical economic research.

MIMA Search

授業計画

1. Introduction and Regression 2. Tree Algorithms and Ensemble Learning 3. Neural Network Learning (1): Feedforward Neural Networks, Training Methods 4. Neural Network Learning (2): Convolutional Neural Networks, Image Data Analysis 5. Neural Network Learning (3): Recurrent Neural Networks, Sequential Data Analysis (Time Series and Text) 6. Machine Learning Applications to Causal Inference (1): Causal Forests 7. Machine Learning Applications to Causal Inference (2): Double/Debiased Machine Learning 8. Machine Learning Applications to Causal Inference (3): Policy Learning 9. Bandit Algorithms 10–13. Student Presentations The lecture plan is subject to change.

授業の方法

Lectures, demonstration programs (Python or R), and students presentations

成績評価方法

The course evaluation is based on two or three problem sets (30%), a student presentation (35%), and a term paper (35%).

履修上の注意

- Students are expected to have a basic knowledge of probability and statistics. - Programming experience is helpful but not required. - Prior knowledge of statistical causal inference is helpful but not required.