Invariance and Causality

Abstract: Why are we interested in the causal structure of a process? In classical prediction tasks, for example, it seems that no causal knowledge is required. In many situations, however, we are interested in a system’s behavior after parts of this system have been changed. Here, causal models become important because they are usually considered invariant under those changes. A causal prediction (which uses only direct causes of the target variable as predictors) remains valid even if we intervene on predictor variables or change the whole experimental setting. We show how we can use invariance in order to estimate the causal structure. This talk does not require any knowledge about causal concepts.

Biography: Jonas is an associate professor in statistics at the University of Copenhagen, and he is a member of the “Junge Akademie”. Previously, Jonas has been leading the causality group at the MPI for Intelligent Systems in Tuebingen and was a Marie Curie fellow at the Seminar for Statistics, ETH Zurich. He studied Mathematics in Heidelberg and Cambridge and did his PhD with B. Schoelkopf, D. Janzing and P. Buehlmann, his thesis received the ETH medal. He is interested in inferring causal relationships from observational data and works both on theory and methodology. His work relates to areas as computational statistics, graphical models, independence testing or high-dimensional statistics.

There will be an additional lectures “Four Lectures on Causality” scheduled on Wednesday 5/10 (10am-12pm & 2pm-4pm) and Thursday 5/11 (10am-12pm & 2pm-4pm) in E18-304.