Causal Artificial Intelligence for Healthcare

3 units

This course provides an introduction to causal artificial intelligence for healthcare, focusing on how causal reasoning and modern machine learning methods can be used to support reliable decision-making from real-world health data. Students will learn how to move beyond prediction toward answering clinically meaningful “what if” questions, such as estimating treatment effects, understanding sources of bias, and evaluating model robustness across populations and settings. Topics include randomized and observational study designs, causal diagrams, target trial emulation, counterfactual reasoning, heterogeneous treatment effects, invariance and transportability, algorithmic fairness, and emerging approaches to causal representation learning. Throughout the course, methodological concepts are motivated by real clinical and public-health use cases.