Reasoning about Independence in Probabilistic Models of Relational Data
This addresses the issue of handling non-i.i.d. data in relational models for researchers in probabilistic machine learning, representing an incremental extension of existing theory.
The paper tackled the problem of inaccurately inferring conditional independence in probabilistic models of relational data using standard d-separation, and introduced relational d-separation with an abstract ground graph representation, resulting in a sound, complete, and computationally efficient method as demonstrated empirically.
We extend the theory of d-separation to cases in which data instances are not independent and identically distributed. We show that applying the rules of d-separation directly to the structure of probabilistic models of relational data inaccurately infers conditional independence. We introduce relational d-separation, a theory for deriving conditional independence facts from relational models. We provide a new representation, the abstract ground graph, that enables a sound, complete, and computationally efficient method for answering d-separation queries about relational models, and we present empirical results that demonstrate effectiveness.