Causal Structure
Causal structure research aims to uncover cause-and-effect relationships within complex systems, using both observational and interventional data to build accurate causal models. Current research focuses on developing robust algorithms for causal discovery, including those leveraging graph neural networks, score-matching techniques, and reinforcement learning to learn directed acyclic graphs (DAGs) efficiently, even in the presence of latent variables and non-linear relationships. These advancements are crucial for improving the reliability and interpretability of machine learning models across diverse fields, from healthcare and robotics to climate science and marketing, enabling more effective decision-making based on a deeper understanding of causality.
Papers
Learning Causal Overhypotheses through Exploration in Children and Computational Models
Eliza Kosoy, Adrian Liu, Jasmine Collins, David M Chan, Jessica B Hamrick, Nan Rosemary Ke, Sandy H Huang, Bryanna Kaufmann, John Canny, Alison Gopnik
Diffusion Causal Models for Counterfactual Estimation
Pedro Sanchez, Sotirios A. Tsaftaris
Disentangled Counterfactual Recurrent Networks for Treatment Effect Inference over Time
Jeroen Berrevoets, Alicia Curth, Ioana Bica, Eoin McKinney, Mihaela van der Schaar
Tell me why! Explanations support learning relational and causal structure
Andrew K. Lampinen, Nicholas A. Roy, Ishita Dasgupta, Stephanie C. Y. Chan, Allison C. Tam, James L. McClelland, Chen Yan, Adam Santoro, Neil C. Rabinowitz, Jane X. Wang, Felix Hill