Functionally compatible local characteristics for the local specification of priors in graphical models

被引:8
作者
Asci, Claudio [1 ]
Piccioni, Mauro [1 ]
机构
[1] Univ Roma La Sapienza, Dipartimento Matemat, I-00185 Rome, Italy
关键词
Gibbs sampler; graphical models; marginal problem;
D O I
10.1111/j.1467-9469.2006.00551.x
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 ; 070103 ; 0714 ;
摘要
The local specification of priors in non-decomposable graphical models does not necessarily yield a proper joint prior for all the parameters of the model. Using results concerning general exponential families with cuts, we derive specific results for the multivariate Gamma distribution (conjugate prior for Poisson counts) and the Wishart distribution (conjugate prior for Gaussian models). These results link the existence of a locally specified joint prior to the solvability of a related marginal problem over the cliques of the graph.
引用
收藏
页码:829 / 840
页数:12
相关论文
共 12 条
  • [1] Asci C, 2003, KYBERNETIKA, V39, P731
  • [2] BARNDORFFNIELSE.OE, 1978, INFORM EXPONENTIAL F
  • [3] Reference priors for discrete graphical models
    Consonni, G
    Leucari, V
    [J]. BIOMETRIKA, 2006, 93 (01) : 23 - 40
  • [4] HYPER MARKOV LAWS IN THE STATISTICAL-ANALYSIS OF DECOMPOSABLE GRAPHICAL MODELS
    DAWID, AP
    LAURITZEN, SL
    [J]. ANNALS OF STATISTICS, 1993, 21 (03) : 1272 - 1317
  • [5] CONJUGATE PRIORS FOR EXPONENTIAL FAMILIES
    DIACONIS, P
    YLVISAKER, D
    [J]. ANNALS OF STATISTICS, 1979, 7 (02) : 269 - 281
  • [6] Gelman A, 2013, BAYESIAN DATA ANAL, DOI DOI 10.1201/9780429258411
  • [7] Reference priors for exponential families
    Gutiérrez-Peña, E
    Rueda, R
    [J]. JOURNAL OF STATISTICAL PLANNING AND INFERENCE, 2003, 110 (1-2) : 35 - 54
  • [8] Functional compatibility, Markov chains, and Gibbs sampling with improper posteriors
    Hobert, JP
    Casella, G
    [J]. JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS, 1998, 7 (01) : 42 - 60
  • [9] JIROUSEK R, 1991, KYBERNETIKA, V27, P403
  • [10] Lauritzen S. L., 1996, GRAPH MODELS