In this paper we compare two alternative MCMC samplers for the Bayesian analysis of
discrete graphical models; we present both a hierarchical and a nonhierarchical
version of them. We first consider the MC3 algorithm by Madigan and York (1995) for which we propose an extension
that allows for a hierarchical prior on the cell counts. We then describe a novel
methodology based on a reversible jump sampler. As a prior distribution we assign,
for each given graph, a hyper-Dirichlet distribution on the matrix of cell
probabilities. Two applications to real data are presented.
Barbieri MM
,
Berger J
(2004)
Optimal predictive model selection
. The Annals of Statistics, 32(3), in press.
2.
Brooks SP
,
Giudici P
,
Roberts GO
(2003)
Efficient construction of reversible jump MCMC proposal distributions, with discussion
. Journal of The Royal Statistical Society, Series B,
65,
3-55
.
3.
Dawid AP
,
Lauritzen SL
(1993)
Hyper Markov laws in the statistical analysis of decomposable graphical models
. The Annals of Statistics, 21,
1272-1317
.
4.
Dawid P
,
Lauritzen SL
(2000) Compatible prior distributions. In
George E
,
Nanopolous P
eds. Bayesian methods with applications to science, policy, and official
statistics.
Luxemburg: Eurostat
, 109-118.
5.
Dellaportas P
,
Forster JJ
(1999)
Markov chain Monte Carlo model determination for hierarchical and graphical
log-linear models
. Biometrika, 86,
615-633
.
6.
Fahrmeir L
,
Tutz G
(1994). Multivariate statistical modelling based on
generalized linear models.
New York: Springer
.
7.
Fowlkes EB
,
Freeny AE
,
Landwehr JM
(1988)
Evaluating logistic models for large contingency tables
. Journal of The American Statistical Association,
83,
611-622
.
8.
Frydenberg M
,
Lauritzen SL
(1989)
Decomposition of maximum likelihood in mixed interaction models
. Biometrika, 76,
539-555
.
9.
Giudici P
,
Green PJ
(1999)
Decomposable graphical Gaussian model determination
. Biometrika, 86,
785-801
.
10.
Giudici P
,
Tarantola C
(1996)
Global prior distribution for discrete graphical models
. Journal of the Italian Statistical Society, 5,
129-147
.
11.
Giudici P
(2000)
Bayesian data mining, with application to benchmarching and credit scoring
. Applied Stochastic Models in Business and Industry,
16,
69-81
.
12.
Giudici P
,
Castelo R
(2003)
Improving Markov chain Monte Carlo search for data mining
. Machine Learning, 50,
127-158
.
13.
Green PJ
(1995)
Reversible jump Markov chain Monte Carlo computation and Bayesian model determination
. Biometrika, 82,
711-732
.
14.
Green PJ
(2003) Trans-dimensional Markov chain Monte
Carlo. In
Green PJ
,
Hjort N
,
Richardson S
eds. Highly structured stochastic systems,
Oxford: Oxford University Press
, 179-196.
15.
Hand DJ
,
Henley WE
(1997)
Statistical classification methods in consumer credit scoring: a review
. Journal of The Royal Statistical Society, Series A,
160,
523-541
.
16.
Hand DJ
,
McConway KJ
,
Stanghellini E
(1997)
Graphical models of applicants for credit
. Journal of Mathematics Applied in Business and Industry,
8,
143-155
.
17.
King R
,
Brooks SP
(2001a)
Prior induction in log-linear models for general contingency table analysis
. The Annals of Statistics, 29,
715-747
.
18.
King R
,
Brooks SP
(2001b)
On the Bayesian analysis of population size
. Biometrika, 88,
317-336
.
19.
Jensen FV
(1996) An introduction to Bayesian networks.
London: University College, London Press
.
20.
Lauritzen SL
(1996) Graphical models.
Oxford: Oxford University Press
.
21.
Madigan D
,
Raftery AE
(1994)
Model selection and accounting for model uncertainty in graphical models
using Occam’s window
. Journal of the American Statistical Association,
89,
1535-1546
.
22.
Madigan D
,
York J
(1995)
Bayesian graphical models for discrete data
. International Statistical Review, 63,
215-232
.
23.
Madigan D
,
York J
(1997)
Bayesian methods for the estimation of the size of a closed population
. Biometrika, 84,
19-31
.
24.
Robert CP
(1994) The Bayesian choice.
New York: Springer-Verlag
.
25.
Tarjan RE
,
Yannakakis M
(1984)
Simple linear-time algorithms to test chordality of graphs, test acyclicity
of hypergraphs, and selectivity reduce acyclic hypergraphs
. SIAM Journal of Computing, 13,
566-579
.
26.
Wormald N
(1985)
Counting labelled chordal graphs
. Graphs and Combinatorics, 1,
193-200
.
