Information Flows in Causal Networks
Nihat Ay and Daniel Polani
Contact the author: Please use for correspondence this email.
Submission date: 02. May. 2006
published in: Advances in complex systems, 11 (2008) 1, p. 17-41
DOI number (of the published article): 10.1142/S0219525908001465
MSC-Numbers: 94A15, 94A40, 94B15
Keywords and phrases: information flow, causality, directed acyclic graphs, mutual information
We introduce a notion of causal independence based on virtual intervention, which is a fundamental concept of the theory of causal networks. Causal independence allows for defining a measure for the strength of a causal effect. We call this information flow and compare it with known information flow measures such as the transfer entropy.