Deleted text in red / Inserted text in green

WW

HEADERS_END

Bayes' Theorem says that if event /B/ has a non-zero

probability of occuring, then when it does occur,

we can compute the probability that event /A/ also

happened as follows:

* EQN:P(A|B)=\frac{P(B|A).P(A)}{P(B)}

This gives a sort of "inverse probability" calculation.

See also:

* http://en.wikipedia.org/wiki/Bayes'_theorem

A recent use of this theorem is spam filtering. Although the application is not strictly

correct, it's close enough that it can be very effective. An alternative, or even

supplementary spam filtering system uses Markov Chains.