Discriminating between causal structures in bayesian networks given partial observations

Publikationstyp
Journal Article
Date Issued
2014
Sprache
English
Author(s)
Moritz, Philipp  
Reichardt, Jörg  
Ay, Nihat  
TORE-URI
http://hdl.handle.net/11420/14347
Journal
Kybernetika  
Volume
50
Issue
2
Start Page
284
End Page
295
Citation
Kybernetika 50 (2): 284-295 (2014)
Publisher DOI
10.14736/kyb-2014-2-0284
Scopus ID
2-s2.0-84901447118
Given a fixed dependency graph G that describes a Bayesian network of binary variables X1, Xn, our main result is a tight bound on P the mutual information Ic(Y1, Yk) Σkj=1 H(Yj)/c - H(Y1, Yk) of an observed subset Y1, Yk of the variables X1, Xn. Our bound depends on certain quantities that can be computed from the connective structure of the nodes in G. Thus it allows to discriminate between different dependency graphs for a probability distribution, as we show from numerical experiments.
Subjects
Bayesian networks
Causal inference
Causal Markov condition
Common ancestors
Information inequalities
Information theory
DDC Class
004: Informatik