Dear Jeremy,
This case is not a simple one.
Much depends on A,B,C dependancy.
If the dependancy is defined in pairs only, there is no way to calculate
P(A|C) having P(A|B) and P(B|C).
You need to have complete 3-dimensional simultaneous distribution of A,B,C.
In such a case you will be able to write analytical expressions of
*any* conditional probabilities you need.
(see any Probability Theory textbook on n-dimensional distribution
formulas).
Sincerely yours,
Dr. Fairman
==Advanced help for students in doing assignments on ANY subjects==
http://expert2002.50megs.com/consultant.html
[EMAIL PROTECTED] (Jeremy Caron) wrote in message
news:<[EMAIL PROTECTED]>...
> Hello,
>
> I was wondering the following:
>
> Given P(A|B) and P(B|C), how strong of an estimate can I make on
> P(A|C) and how do I go about doing so?
>
> Thanks in advance for any help!
>
> - Jeremy
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
. http://jse.stat.ncsu.edu/ .
=================================================================
