Causality in Bayesian Belief Networks

TL;DR

This paper explores how Bayesian belief networks can be given causal interpretations by linking them to structural equation models, emphasizing mechanisms and causal asymmetries.

Contribution

It establishes conditions under which BBNs can be causally interpreted, bridging the gap between BBNs and structural equation models.

Findings

Causality in BBNs can be understood through mechanisms.

Conditions for causal interpretation of BBNs are formulated.

Mechanism-based causality applies to BBNs when certain criteria are met.

Abstract

We address the problem of causal interpretation of the graphical structure of Bayesian belief networks (BBNs). We review the concept of causality explicated in the domain of structural equations models and show that it is applicable to BBNs. In this view, which we call mechanism-based, causality is defined within models and causal asymmetries arise when mechanisms are placed in the context of a system. We lay the link between structural equations models and BBNs models and formulate the conditions under which the latter can be given causal interpretation.