Lucas Kook 

Date/Time: 02.04.2025, 12:30 

Location: D2.2.094 

Abstract 

In many scientific disciplines, experiments are prohibitively expensive or unethical to conduct. Therefore, researchers need to rely on non-experimental (observational) data and strong assumptions to draw causal conclusions. In this talk, I will discuss (i) structural causal models as a statistical framework for causal inference, (ii) under what assumptions and how causal structures can be learned from observational data, and (iii) challenges that arise when dealing with complex data structures, such as images, text, or dependent data. Finally, I will discuss what role causality can play within bilateral artificial intelligence. 

Bio 

https://lucaskook.github.io/