## can you help me to build the Beyesian net by using GeNIe

The front end.
youyou
Posts: 1
Joined: Thu Nov 25, 2010 7:54 am

### can you help me to build the Beyesian net by using GeNIe

I am a new learner in GeNIe. I have been troubled by Bayesian problem.

The problem is stated as follows:

there are three experts to evaluate the security policy of car (s),

A. the first expert (E1) gives three rules:

1) if the security policy of car is "poor", the probability of the accident is 0.4, i.e.,
P(S|E11)=0.4.
2) if the security policy of car is "middle", the probability of the accident is 0.2, i.e.,
P(S|E12)=0.2.
3) if the security policy of car is "good", the probability of the accident is 0.1, i.e.,
P(S|E13)=0.1.

So, we can get the probability of the security policy locted in every rule when the accident existed:
P(E11|S=yes)=0.5714, P(E12|S=yes)=0.2857, P(E13|S=yes)=0.1429.
Correspondingly, the probability of the security policy locted in every rule when the accident not existed:
P(E11|S=No)=0.2609, P(E12|S=no)=0.3478, P(E13|S=no)=0.3913.

B. the second expert (E2) gives three rules:

1) if the security policy of car is "poor", the probability of the accident is 0.5, i.e.,
P(S|E21)=0.5.
2) if the security policy of car is "middle", the probability of the accident is 0.2, i.e.,
P(S|E22)=0.2.
3) if the security policy of car is "good", the probability of the accident is 0.1, i.e.,
P(S|E23)=0.1.

So, we can get the probability of the security policy locted in every rule when the accident existed:
P(E21|S=yes)=0.625, P(E22|S=yes)=0.25, P(E23|S=yes)=0.125.
Correspondingly, the probability of the security policy locted in every rule when the accident not existed:
P(E21|S=No)=0.2272, P(E22|S=no)=0.3636, P(E23|S=no)=0.4092.

C. the third expert (E3) gives three rules:

1) if the security policy of car is "poor", the probability of the accident is 0.4, i.e.,
P(S|E31)=0.4.
2) if the security policy of car is "middle", the probability of the accident is 0.2, i.e.,
P(S|E32)=0.3.
3) if the security policy of car is "good", the probability of the accident is 0.1, i.e.,
P(S|E33)=0.1.

So, we can get the probability of the security policy locted in every rule when the accident existed:
P(E31|S=yes)=0.5, P(E32|S=yes)=0.375, P(E33|S=yes)=0.125.
Correspondingly, the probability of the security policy locted in every rule when the accident not existed: P(E31|S=No)=0.2727, P(E32|S=no)=0.3182, p
(E33|S=no)=0.4091.

Based on above procedures, I donot cumputing the the probability of the accident, i.e., p(S)=? by GeNIe. The attarched file can not get p(S).
Can you help me to build the Beyesian net by using GeNIe at your free.
Attachments
Network2.xdsl
my GeNIe file