Ciarán Gilligan-Lee -- Learning and reasoning at different levels of Pearl's Causal Hierarchy

Learning and reasoning at different levels of Pearl's Causal Hierarchy
Dr Ciarán Gilligan-Lee, University College London

Abstract: Causal reasoning is vital for effective reasoning in many domains, from healthcare to economics. In medical diagnosis, for example, a doctor aims to explain a patient’s symptoms by determining the diseases causing them. This is because causal relations, unlike correlations, allow one to reason about the consequences of possible treatments and to answer counterfactual queries. In this talk I will present two recent causal inference projects done with my collaborators. One of which is concerned about the ability to disentangle the effect of multiple treatments in the presence of hidden confounders. The other is about how one can learn and reason with counterfactual distributions. In both cases I will strive to motivate and contextualise the results with real word examples. Along the way I'll discuss the distinction between learning and inference in the context of a causal model. The talk is based on the following two papers: https://arxiv.org/pdf/2109.01904.pdf and http://adrem.uantwerpen.be/bibrem/pub....

Speaker bio: Dr. Ciarán Gilligan-Lee is Head of the Causal Inference research lab at Spotify and an Honorary Associate Professor at University College London. Previously he was a Senior Researcher at Babylon Health, where he built & led the Causal Machine Learning team. His research has been presented at some of the top AI conferences, such as AAAI and NeurIPS, and has been covered by popular media outlets such as MIT Technology Review, New Scientist, The Times, The Independent, Newsweek, and Gizmodo. Before joining industry, he was PI of an EPSRC grant at University College London on causal inference and its applications. He holds a DPhil (PhD) from University of Oxford Department of Computer Science.

http://www.educationalneuroscience.or...