Hi,
I am working on Bayesian updating of probabilities in large Bayesian
networks containing a lot of continuous nodes with non-conjugate
probability densities and sparse observations. I�ve used WinBugs,
which is rather good in doing (approximate) inference on such networks
but I�ve been unable to find a programming-oriented description of
the algorithms they are using (I�m aware of Gilks & al., �Markov
Chain Monte Carlo in Practice�).
Here are my questions: Could anybody point out references on �ready
to implement� algorithms (rather than theoretical analyses) on
Gibbs sampling for hybrid and/or continuous DAGs? Are there
alternatives methods (theory, algorithms or software) to Gibbs
sampling for this kind of problem?
Thank you in advance.
Philippe
__________________________________________________
Philippe WEBER Associate Professor
CRAN - CNRS UMR 7039
Universit� Henri Poincar�, Nancy 1 - ESSTIN
2, rue Jean Lamour
54519 VANDOEUVRE-LES-NANCY Cedex (France)
E-Mail : [EMAIL PROTECTED]
t�l +33 (0)3 83 68 51 27 fax +33 (0)3 83 68 50 12
__________________________________________________
