I'm working on just that. My thought, and I must admit I don't know all the theory involved, is that the effectiveness of Bayes and the FN/FP is purely a function of how you trained it. The *ONLY* way (in my mind) to score the Bayes functions would be to train it and run the Mass-check. I'm working on that process right now.
I also added some more granularity of scoring at both ends of the % scale, since confidence should be much higher as pass 90% and approach 99%? It makes sense to me, but who knows. When I get the GA running, I will have some facts and not just thoughts.... <<Dan>> | -----Original Message----- | From: Bob Apthorpe [mailto:[EMAIL PROTECTED] | Sent: Thursday, November 20, 2003 3:34 PM | To: SATalk list | Short answer: all scores including those from Bayes are | generated by a genetic algorithm ("the GA") which cares | little for making the scores fit an ideal curve (normal | distribution) or satisfy the consistency hobgoblins. The GA | adjusts scores until FNs and FPs are minimized within the | time and accuracy constraints it's given. ------------------------------------------------------- This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/ _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
