Over the past decade, the use of unstructured data has been growing steadily in economics and related disciplines, with a rapid acceleration in the wake of COVID-19. This Machine Learning for Textual and Unstructured Data Seminar, organized by IESE and the Fundación Ramón Areces, is aimed at introducing machine learning methods for the analysis of text and unstructured data to an audience of academics and researchers in the fields of finance, economics, and accounting.

An outline of the course structure and objectives:

• This course will begin with an overview of the challenges and opportunities of working with such data with a focus on natural language.
• We first review relatively straightforward methods that operate on raw word counts across documents before studying machine learning algorithms for dimensionality reduction, which is a key problem in the analysis of text and unstructured data. These encompass factor models whose basic structure is similar to well-known econometric methods, as well as neural network models which form the basis of much of modern natural language processing.
• We also show how these ideas can be applied to non-textual data such as surveys, images and scanner data on goods purchases.
• Coding will be done in Python. Knowledge of Python is not required, and we will provide extra materials and in-class support on coding for participants.
• Finally, we will discuss how these methods are employed in empirical research in economics, accounting and finance.
• Participants will also have an opportunity to present their own research in an informal but constructive environment at the end of each day.

The seminar forms part of IESE’s Center for International Finance (CIF), which advocates a multidisciplinary approach to generate and disseminate knowledge and promotes best practices in financial institutions and capital markets.