hi,
probably a good point to start with is generating
dags randomly:
G. Melancon, I. Dutour and M. Bousquet-Melou
Random generation of dags for graph drawing.
Technical Report INS-R0005 February, 2000.
Dutch Research Center for Mathematics and Computer Science (CWI).
ISSN 1386-3681
http://www.cwi.nl/ftp/CWIreports/INS/INS-R0005.ps.Z
they are generated by constructing a Markov chain, so by
limiting the number of parents during the process, you can
probably obtain dags that resemble those used as Bayesian
Networks from real-world domains.
once you get your dag at random, you can generate your cpt's
for each variable separately (because of global independence)
and in the way you prefer.
best,
robert.
On Fri, 26 Oct 2001, Fabio Gagliardi Cozman wrote:
> Date: Fri, 26 Oct 2001 08:05:58 -0700
> To: uai@cs.orst.edu
> From: Fabio Gagliardi Cozman <fgcozman@usp.br>
> Sender: owner-uai@cs.orst.edu
> Subject: [UAI] Generating Bayes nets randomly
>
> Dear UAIers,
>
> I would appreciate if anyone could give some pointers
> on how to generate Bayesian networks randomly. That is,
> how to construct directed acyclic graphs and their
> associated parameters in a way that uniformly samples
> the space of the DAGs? I have not found a simple explanation
> on how to do it, even though many papers refer to networks
> that have been generated randomly. I wonder if there is
> software available for this.
>
> Maybe the problem is simple, but it does seem to require
> some sophistication. Even to generate the conditional
> probabilities, it does not seem that simply generating
> tuples and normalizing them will uniformly sample the
> space of probability distributions.
>
> Any help is appreciated. Thanks,
>
> Fabio Cozman
>
>
This archive was generated by hypermail 2b29 : Mon Oct 29 2001 - 09:57:34 PST