Causal Bayesian Experimental Design and Representation Learning

Please note: this event is in hybrid format, taking place in Science Park room L3.36 and streamed via Zoom: https://uva-live.zoom.us/j/6466222109

Abstract: In the first part of my talk, I will present Active Bayesian Causal Inference (ABCI), a fully-Bayesian active learning framework for integrated causal discovery and reasoning, which jointly infers a posterior over causal models and queries of interest. In our approach to ABCI, we focus on the class of causally-sufficient, nonlinear additive noise models, which we model using Gaussian processes. We sequentially design experiments that are maximally informative about our target causal query, collect the corresponding interventional data, and update our beliefs to choose the next experiment.
In the second part of the talk, I will give a brief overview of some of our recent work on identifiable (causal) representation learning.

Bio: Julius von Kügelgen is a 5th year PhD student in the Cambridge-Tübingen program, supervised by Bernhard Schölkopf at the Max Planck Institute for Intelligent Systems and by Adrian Weller at the University of Cambridge. His research interests lie at the intersection of causal inference and machine learning, including causal reasoning for explainability, recourse and fairness; causal discovery; and identifiable causal representation learning. He has interned at Amazon (2019 - 2021) and was awarded a Google PhD Fellowship in Machine Learning in 2022. Previously, he studied Mathematics (BSc+MSci) at Imperial College London and Artificial Intelligence (MSc) at UPC Barcelona in Spain and at TU Delft in the Netherlands.