Robert Dodier wrote:
>
> Julie asks,
>
> > Is there any algorithms of Bayesian Network to work directly on the
> > mixture of continuous and categorical variables?
> [...]
>
> I know of a few possible approaches. One is to discretize all the
> continuous variables. Another is to approximate the distributions
> of the continuous variables by conditional Gaussians; exact
> algorithms are known for loopy Bayesian networks with discrete and
> conditional Gaussian continuous variables, although IIRC no
> discrete variable can be a child of a continuous variable.
I have seen this restriction that no discrete child can be a child of a
continuous variable written down as if it can't be done. Why? A
disceteization is exactly what a discrete child of a continuous variable
is. I would expect that making the discretization explicit would be
advantageous, as, for example, different discretizations may be
appropriate for different purposes.
Has anyone looked at finding optimal discretizations by having a
discrete variable as a child of a continuous variable and then
optimizing over the distribution? [I was teaching my class about mixing
continuous and discrete variables and sketched out how this could be
done, but I couldn't find a reference.]
David
P.S. What does "IIRC" mean?
This archive was generated by hypermail 2b29 : Fri Apr 07 2000 - 11:31:48 PDT