Bayes theorem is a principled way to calculate posterior
probability of each hypothesis
Assumptions: The training data is noise free (i.e., ). The target
concept is contained in . We have no a priori reason to believe
that any hypothesis is more probable than any other.
for all in
So if is inconsistent with ,
So if is consistent with ,
every consistent hypothesis is a MAP hypothesis
Patricia Jean Riddle
Wed Jun 23 13:06:34 NZST 1999