LFU:online

Institut für Wirtschaftstheorie, -politik und -geschichte Institut für Wirtschaftstheorie, -politik und -geschichte

Ass.-Prof. Dr. Marica Valente Ass.-Prof. Dr. Marica Valente, +43 512 507 71026

432125

VU Wirtschaftspolitik: Methoden der Datenanalyse in der Umweltökonomie

VU 3

7,5

wöch.

jährlich

Englisch

This course provides a broad introduction to microeconometric empirical methods for economists, including traditional econometric methods and machine learning techniques. The target audience are master students interested in learning how to perform data analysis, outcome prediction and policy evaluation. Students will learn how to use the statistical software R. Applications will focus on environmental policies for climate change mitigation (waste and air pollution control). Completing the course will enable students to conduct independent empirical research in their master thesis as well as future jobs (e.g. public policy institutions, consulting firms, and doctoral programs).

Machine learning (ML) defines a set of modern empirical tools used in fields like statistics, computer science, AI and, more recently, economics. ML in economics is often viewed as a black-box: this course aims to make ML less obscure and more accessible. In this course, we will walk through the basics of ML with a focus on supervised learning such as regularized linear regression and tree-based methods for both prediction and causal effect estimation. In addition, I will show R codes to familiarize with the algorithms’ implementation. Existing statistical packages make it trivial to do ML in practice. However, I will show how economic intuition still plays a crucial role in improving the algorithms’ performance. At the end of the course, students will know how to use ML methods to solve problems that traditional econometrics cannot.

1. Statistics, econometrics and machine learning.

2. Draw contrasts with traditional approaches (OLS in classical statistics)

3. How to use machine learning methods for prediction?

4. How to use machine learning tools in R?

5. Tree-based methods in R. Explain homework assignment

6. Parametric methods. Regression-Based Methods: Lasso (Ridge, Bridge, and Elastic Nets)

7. Analyze regression-based methods in R

8. Increasing prediction accuracy: Superlearning / Stacking / Ensemble Selection from Libraries of Models

9. How to conduct empirical research. How to present research idea / description of available high-dimensional data sources

10. How to write an empirical paper?

11. Causal methods for policy evaluation: Traditional econometric methods

12. Causal methods for policy evaluation: Machine learning methods

13. Student oral presentation of research ideas + Feedback

siehe Termine