Reading Dependencies from Covariance Graphs

Abstract

The covariance graph (aka bi-directed graph) of a probability distribution $p$ is the undirected graph $G$ where two nodes are adjacent iff their corresponding random variables are marginally dependent in $p$. In this paper, we present a graphical criterion for reading dependencies from $G$, under the assumption that $p$ satisfies the graphoid properties as well as weak transitivity and composition. We prove that the graphical criterion is sound and complete in certain sense. We argue that our assumptions are not too restrictive. For instance, all the regular Gaussian probability distributions satisfy them.

Changes from v1 to v2: Minor cosmetic changes, plus the addition of reference (Richardson and Spirtes, 2002) in page 8. Changes from v2 to v3: Addition of some references; International Journal of Approximate Reasoning, 2012