Latent-free equivalent mDAGs

Abstract

We show that the marginal model for a discrete directed acyclic graph (DAG) with hidden variables is distributionally equivalent to another fully observable DAG model if and only if it does not induce any non-trivial inequality constraints.

13 pages, 6 figures

References in corpus (4)

• Beyond Bell's Theorem: Correlation Scenarios
• The Inflation Technique for Causal Inference with Latent Variables
• Bell inequalities for continuous-variable correlations
• Nested Markov Properties for Acyclic Directed Mixed Graphs