Bayesian Causal Discovery

Bayesian Causal Discovery (BCD) aims to infer causal relationships between variables by probabilistically representing uncertainty in the underlying causal structure, typically represented as a directed acyclic graph (DAG). Current research focuses on developing scalable algorithms, such as variational inference and Markov Chain Monte Carlo methods, to efficiently explore the vast space of possible DAGs, often employing continuous relaxations to overcome computational challenges associated with the acyclicity constraint. This work is crucial for improving the reliability and accuracy of causal inference across diverse fields, from neuroscience and genetics to autonomous systems, where robust causal models are essential for decision-making and understanding complex systems. A key challenge remains the development of robust evaluation metrics for comparing different BCD methods, particularly in low-data scenarios.

Papers

July 6, 2024

Scalable Variational Causal Discovery Unconstrained by Acyclicity
Nu Hoang, Bao Duong, Thin Nguyen
Variational Inference Causal Graph Directed Acyclic Graph Bayesian Causal Discovery Causal Theory

June 5, 2024

Challenges and Considerations in the Evaluation of Bayesian Causal Discovery
Amir Mohammad Karimi Mamaghan, Panagiotis Tigas, Karl Henrik Johansson, Yarin Gal, Yashas Annadani, Stefan Bauer
Technical Challenge Global Evaluation Causal Discovery Causal Graph Causal Model Crucial Consideration Bayesian Causal Discovery

December 1, 2023

Bayesian causal discovery from unknown general interventions
Alessandro Mascaro, Federico Castelletti
Causal Discovery Bayesian Inference Directed Acyclic Graph Markov Equivalence Class Interventional Data Bayesian Causal Discovery Unknown Intervention

August 19, 2023

Towards Probabilistic Causal Discovery, Inference & Explanations for Autonomous Drones in Mine Surveying Tasks
Ricardo Cannizzaro, Rhys Howard, Paulina Lewinska, Lars Kunze
Line by Line Explanation Scientific Inference Causal Model Causal Pattern Autonomous Drone Novel Causal Model Bayesian Causal Discovery

July 26, 2023

BayesDAG: Gradient-Based Posterior Inference for Causal Discovery
Yashas Annadani, Nick Pawlowski, Joel Jennings, Stefan Bauer, Cheng Zhang, Wenbo Gong
Causal Discovery Posterior Inference DAG Learning Bayesian Causal Discovery

July 11, 2023

Benchmarking Bayesian Causal Discovery Methods for Downstream Treatment Effect Estimation
Chris Chinenye Emezue, Alexandre Drouin, Tristan Deleu, Stefan Bauer, Yoshua Bengio
Causal Inference Causal Discovery Treatment Effect Causal Discovery Method Bayesian Causal Discovery

February 10, 2023

Brain Effective Connectome based on fMRI and DTI Data: Bayesian Causal Learning and Assessment
Abdolmahdi Bagheri, Mahdi Dehshiri, Yamin Bagheri, Alireza Akhondi-Asl, Babak Nadjar Araabi
Causal Discovery Functional Magnetic Resonance Imaging Direct Assessment Human Connectome Project Bayesian Causal Discovery Effective Connectome

July 12, 2022

Latent Variable Models for Bayesian Causal Discovery
Jithendaraa Subramanian, Yashas Annadani, Ivaxi Sheth, Stefan Bauer, Derek Nowrouzezahrai, Samira Ebrahimi Kahou
Causal Discovery Latent Variable Model Causal Representation Bayesian Causal Discovery

March 3, 2022

Interventions, Where and How? Experimental Design for Causal Models at Scale
Panagiotis Tigas, Yashas Annadani, Andrew Jesson, Bernhard Schölkopf, Yarin Gal, Stefan Bauer
Causal Discovery Visual Analogue Scale Structural Causal Model Experimental Design Health Intervention Bayesian Causal Discovery Active Causal

January 28, 2022

BCDAG: An R package for Bayesian structure and Causal learning of Gaussian DAGs
Federico Castelletti, Alessandro Mascaro
Causal Discovery Causal Relationship Causal Effect Estimation Causal Learning Causal Structure Learning R Package Bayesian Structure Bayesian Causal Discovery Gaussian DAG

December 6, 2021

BCD Nets: Scalable Variational Approaches for Bayesian Causal Discovery
Chris Cundy, Aditya Grover, Stefano Ermon
Variational Inference Causal Discovery Latent Variable Causal Relationship Scalable Variational Bayesian Causal Discovery