Lecturer: Jaime Sevilla
Time: 17:00 CET
Place: https://timeout.srcf.net/js2545-exb-neg-bsg (access code: 626408)
We will introduce the necessity of causal inference via Simpson’s paradox, the notion of causal graphical models as a way of representing probabilistic causal models, explain how to compute the effect of interventions using a graphical model, and overview the problems of causal discovery (learning the orientation of the edges of a causal model) and causal identifiability (measuring the strength of a causal link).
- Spirtes, Glymour, Scheines Causation, Prediction and Search.
- Pearl, Glymour, P. Jewell Causal Inference in Statistics: A Primer