2021/07/22 by Philippe Brouillard, Perouz Taslakian, Brouillard, Philippe +8 · 3 citations
Computer Science · Mathematics · #Advanced Graph Neural Networks #Algorithm #Artificial intelligence #Bayesian Modeling and Causal Inference #Causal inference #Causal model #Class (philosophy) #Computer science #Directed acyclic graph #Discrete mathematics #Econometrics #Equivalence (formal languages) #Equivalence class (music) #Generative grammar #Graph #Identification (biology) #Machine Learning and Algorithms #Machine learning #Mathematics #Observational equivalence #Set (abstract data type) #Theoretical computer science #cs.AI #cs.LG #stat.ML
paper · pdf · doi:10.48550/arxiv.2107.10703
published in arXiv (Cornell University) (Cornell University) · 30 pages, 13 figures, accepted for the 1st conference on Causal Learning and Reasoning (CLeaR), 2022
openalex publication_date 2021/07/22 · arxiv created 2022/02/28 · arxiv updated 2022/03/01 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/08
Causal discovery from observational data is a challenging task that can only be solved up to a set of equivalent solutions, called an equivalence class. Such classes, which are often large in size, encode uncertainties about the orientation of some edges in the causal graph. In this work, we propose a new set of assumptions that constrain possible causal relationships based on the nature of variables, thus circumscribing the equivalence class. Namely, we introduce typed directed acyclic graphs, in which variable types are used to determine the validity of causal relationships. We demonstrate, both theoretically and empirically, that the proposed assumptions can result in significant gains in the identification of the causal graph. We also propose causal discovery algorithms that make use of these assumptions and demonstrate their benefits on simulated and pseudo-real data.