Economic Machine Learning Lab

Welcome to EML Lab

We are excited to announce the establishment of the Economic Machine Learning (EML) Lab in the Department of Economics at Indiana University. The Lab aims to develop and disseminate innovative methodologies and theories specifically designed for analyzing large-dimensional economic data. In today's data-rich environment, there is an urgent need for tools that can effectively handle complex datasets we encounter in economic applications, and the Lab is poised to meet this demand.

The primary goal of the EML Lab is to equip faculty and students in the IU community with the necessary skills and knowledge to utilize machine learning tools for economic data analysis. The Lab offers a supportive environment where participants can learn various new methodologies while leveraging their theoretical backgrounds. Regular training sessions will be organized to help faculty and students acquire the skills needed to implement standard machine learning tools effectively. Faculty associates and student assistants are also available to provide consultation for those working with large-dimensional data in their research at more advanced levels.

During August 19-23, 2024, the EML Lab hosted its inaugural workshop, Machine Learning in Economics, attended by six faculty members and thirteen graduate students. This workshop included two daily sessions: morning lectures covering the theoretical foundations of various machine learning methodologies and afternoon sessions focused on practical applications in empirical research. The workshop covered a range of topics, including large dimensional regression techniques such as PCA, Ridge, and Lasso; reproducing kernel Hilbert space and kernel methods; support vector machines, including those with kernel tricks; tree-based approaches such as random forests with resampling and boosting; and neural networks in both vanilla and recurrent forms. Participants were encouraged to bring their own laptops and had some working knowledge of programming in MATLAB and Python. A short session on the basics of both programming languages was included to support all attendees.

The EML Lab is committed to fostering collaborative research and sharing ideas among faculty and students to develop new methodologies and theories tailored for economic analysis. We look forward to the contributions and advancements that will emerge from the Lab and our ongoing commitment to excellence in economic research and education.