Martin Spindler - Uni Hamburg

"Inference in High-Dimensional Settings: Theory and Applications"


Abstract

While machine learning methods have been developed for prediction problems in high-dimensional settings, many questions in industry and research are inference questions or causal problems. In this talk we first introduce the so-called Double Machnine Learning (DML) approach which allows for (causal) inference in high-dimensional settings. Next, we apply it to different models, like Gaussian Graphical Models, Additive Models and for Treatment Effect estimation and highlight empirical applications.


Additional information:


Wird geladen