User menu

Accès à distance ? S'identifier sur le proxy UCLouvain

Copula directed acyclic graphs

Primary tabs

download
  • Open access
  • PDF
  • 3.87 M
Pircalabelu, Eugen [UCL] Claeskens, Gerda [KU Leuven] Gijbels, Irène [KU Leuven]

A new methodology for selecting a Bayesian network for continuous data outside the widely used class of multivariate normal distributions is developed. The `copula DAGs' combine directed acyclic graphs and their associated probability models with copula C/D-vines. Bivariate copula densities introduce flexibility in the joint distributions of pairs of nodes in the network. An information criterion is studied for graph selection tailored to the joint modeling of data based on graphs and copulas. Examples and simulation studies show the flexibility and properties of the method.

Document type Article de périodique (Journal article) – Article de recherche
Access type Accès libre
Publication date 2017
Language Anglais
Journal information "Statistics and Computing" - Vol. 27, no. 1, p. 55-78 (2015)
Peer reviewed yes
Publisher Springer Science and Business Media LLC
issn 0960-3174
e-issn 1573-1375
Publication status Publié
Affiliation KU Leuven - ORSTAT
Keywords Theoretical Computer Science ; Statistics ; Probability and Uncertainty ; Statistics and Probability ; Computational Theory and Mathematics
Links
Bibliographic reference Pircalabelu, Eugen ; Claeskens, Gerda ; Gijbels, Irène. Copula directed acyclic graphs. In: Statistics and Computing, Vol. 27, no. 1, p. 55-78 (2015)
Permanent URL http://hdl.handle.net/2078/219723