On 2/28/02 5:28 PM, "Scott Walde" <[EMAIL PROTECTED]> wrote:
> Would it make sense to take a look at the ratio of spam to non-spam for > each given rule, and to constrain the score to either -ve or +ve depending > on which way the ratio leaned? This way, "monsterhut" may wander > randomly, but it will only wander randomly in the +ve direction, or peg > itself at zero. I can't imagine any situation where there were more spams > than non-spams that trigger a rule, yet you want a -ve score. (and > vice-versa) I'm basically planning something like this. The scores are initialized this way now; I'm planning on making the evolution step of the GA tend to move the values of scores in the +ve direction for highly-spammy rules, and tend to move them in the -ve direction for nonspam rules. Basically, when adding gaussian noise to the score, I'll shift the noise either +ve or -ve by some amount based on the ratio of spam/nonspam for that particular rule. C _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
