On Wed, 20 Jan 2016 11:35:33 -0800 Marc Perkel <supp...@junkemailfilter.com> wrote:
> Bayes is about matching. My Evolution filter is about NOT matching. > It's the*NOT matching* that makes it different. Unless you've described it wrong, it's not about not matching. Its about seeing if there are more tokens seen in spam that have never been seen in ham than the opposite. Unless I misunderstand your technique, tokens that have never, ever been seen are not considered at all. [As is indeed the case with Bayes, except for those Bayes classifiers that assign some arbitrary probability to never-before-seen tokens.] So it is, in fact, matching. It looks very much like Bayes to me except without the theoretical justification. Regards, Dianne.