Elias Bareinboim Research Group

Columbia University

How can causal inference methods enable reliable decision-making from biased observational data in real-world systems? The group conducts research on causal inference and probabilistic modeling to build formal methods for reasoning about interventions and counterfactuals. Researchers develop algorithms in sequential decision making that combine probabilistic models with decision-theoretic tools to support robust policy selection. Projects apply causal and probabilistic techniques to domains such as computational biology and computer vision to translate formal causal models into domain-specific analyses. The Causal Artificial Intelligence Lab integrates these methods to study both foundational theory and practical applications in language, vision, and robotics.