Deviance Information Criteria for Missing Data Models

Gilles Celeux (1) , Florence Forbes (1) , Christian P. Robert (1) , Mike Titterington (1)

1 IS2 - Statistical Inference for Industry and Health

The deviance information criterion (DIC) introduced by is directly inspired by linear and generalised linear models, but it is not so naturally defined for missing data models. In this paper, we reassess the criterion for such models, testing the behaviour of various extensions in the cases of mixture and random effect models.

Mots clés: DEVIANCE; COMPLETION; MAP; MIXTURE MODEL; RANDOM EFFECT MODELS

Domaines: Autre [cs.OH]

Gilles Celeux, Florence Forbes, Christian P. Robert, Mike Titterington. Deviance Information Criteria for Missing Data Models. [Research Report] RR-4859, INRIA. 2003. ⟨inria-00071724⟩