Does anyone used the influence diagram functions implemented in BNT ?
I'm studying the examples given in BNT/examples ,
but I have some questions ...
Can we add some evidence in the bnet (on chance nodes or in decision nodes
to force specific actions) before using solve_limid to find the maximum
expected utility and the optimal strategy ?
(i.e. is it possible to compute things like MEU(Decision | Evidence) ?)
I tried on some examples but the evidence added doesn't change the result
:-(
Kevin : Once I will have understood all the functions proposed for utility
and influence diagrams, I will write a simple tutorial in the french BNT
site (based on the oil wildcatter problem), so I could translate it in
english (and you will correct it :-))
Philippe Leray
Associate Professor
Department of Information Systems Engineering - INSA Rouen
Lab PSI (Perception, Systems and Information)
philippe.leray@insa-rouen.fr
http://asi.insa-rouen.fr/~pleray/activities.html
------- End of Forwarded Message
This archive was generated by hypermail 2b29 : Tue Nov 06 2001 - 10:25:38 PST