because the Particle-Filtering algorithm in 15.15 should be applicable in

Dynamic Bayes Networks (DBN), and not only in a DBN with one state variable, I

am trying to implement the algorithm for general DBNs.

I would appreciate any comments to my approach:

I define a DBN as 2 Bayes Networks:

1) the transition model which has parents variables not connected between them,

and some other variables, the Xs variables which can be connected between them

and some have as parents the other ones.

2) the sensor model, with the Xs variables and the Es variables.

For the algorithm 15.15 to work with this approach, only slightly modifications

have to be done. If somebody is interested I can post them.

Maybe somebody could give me some links where I can get simple DBN data to test

my algorithm with.

Regards,

--

Ivan F. Villanueva B.

artificialidea.com

<<< European Community Patent will bring >>>

<<< Software patents by the backdoor >>>

<<< http://wiki.ffii.org/ComPatEn >>>