MIA: Cheng Zhang and Nick Pawlowski, Deep End-to-end Causal Inference; Primer: Causal Discovery

Models Inference and Algorithms December 7, 2022 Broad Institute of MIT and Harvard Meeting: Deep End-to-end Causal Inference Nick Pawlowski Microsoft Causal inference is essential for data-driven decision-making across domains such as business engagement, medical treatment, or policymaking. Building a framework that can answer real-world causal questions at scale is critical. However, research on deep learning, causal discovery, and inference has evolved separately. In this talk, we will present a Deep End-to-end Causal Inference (DECI) framework, a single flow-based method that takes in observational data and can perform both causal discovery and inference, including conditional average treatment effect estimation (CATE). Moreover, we will talk about how such a framework can be used with different real-world data, including time series or considering latent confounders. In the end, we will cover different application scenarios with the Microsoft causal AI suite. We hope that our work bridges the causality and deep learning communities leading to real-world impact. Primer: An Introduction to Causal Discovery and Inference Cheng Zhang Microsoft Causal relationships are fundamental to the way we think about the world und understand it. However, the mathematically formulation of causality has only started relatively recently and is still under-way. This primer will start with a general introduction into the field of causality and the causal hierarchy, mentioning questions associated with all rungs of it. We will briefly introduce structural equation models (SEMs) and Pearl's do-calculus. Furthermore, we will discuss different approaches to identify causal relationships from observational data and introduce different methods for estimating causal effects. For more information visit: https://www.broadinstitute.org/talks/... Copyright Broad Institute, 2022. All rights reserved.