Oct 29
show'n'tell projects
Gibbs Sampling
Homework for Tuesday is posted : one Bayes Network problem,
done two ways : analytic & numeric simulation
I have written up some notes to try to make more explicit
what I think you should know :
... so let's walk through my notes.
Then let's look at this one in class , which is along
the same lines (though more complicated) than your homework:
B = BrokeElectionLaw
I = Indicted
M = PoliticallyMotivatedProsecutor
G = FoundGuilty
J = Jailed.
- The textbook problem : "Calculate the probability that someone goes to jail given that they broke the law, have been indicted, and face a politically motivated prosecutor."
Can we write some code to estimate this with Gibbs Sampling?
How about the looser question P(J|B) ?
Depending on time, I may wave my hands in the direction
of what's in chapter 15 ... but we'll discuss that further next week.
