On intercausal interactions in probabilistic relational models

Silja Rennoij (1) , Linda van Der Gaag (1) , Philippe Leray (2)

1 Universiteit Utrecht
2 LS2N - Laboratoire des Sciences du Numérique de Nantes

Probabilistic relational models (PRMs) extend Bayes-ian networks beyond propositional expressiveness by allowing the representation of multiple interacting classes. For a specific instance of sets of concrete objects per class, a ground Bayesian network is composed by replicating parts of the PRM. The interactions between the objects that are thereby induced, are not always obvious from the PRM. We demonstrate in this paper that the replicative structure of the ground network in fact constrains the space of possible probability distributions and thereby the possible patterns of intercausal interaction.

Mots clés: intercausal interaction; qualitative constraints on probability distributions; PRM instances

Domaines: Apprentissage [cs.LG]

Silja Rennoij, Linda van Der Gaag, Philippe Leray. On intercausal interactions in probabilistic relational models. The Eleventh International Symposium on Imprecise Probability: Theories and Applications (ISIPTA ’19), 2019, Ghent, Belgium. pp.327 - 329. ⟨hal-02129171⟩