EcoSta 2022: Start Registration
View Submission - EcoSta2022
A0545
Title: Modern kernel methods for econometrics Authors:  Krikamol Muandet - Max Planck Institute for Intelligent Systems (Germany) [presenting]
Abstract: While recent developments of kernel methods have led to numerous applications in machine learning and statistics, they have not been fully utilized to solve problems in econometrics. We will provide examples of how modern kernel methods can be employed to solve unique econometric problems ranging from conditional moment (CM) test and distributional treatment effect (DTE) estimation to an instrumental variable (IV) regression. In addition, we will highlight the potential research directions that lie at the intersection of machine learning and economics.