Hello Wolffe & everyone else, on 04-Okt-2005 at 22:45 you (Wolffe) wrote:
> Is there anyway one can reverse the overtraining? Just an idea. You could keep the most recent spam mails, 30 days or so, reset the bayesit database and then train it anew with the spam mails you kept for spam and your sent mail folder for ham. -- Best regards, Alexander (http://www.neurowerx.de - ICQ 238153981) I am not a vegetarian because I love animals; I am a vegetarian because I hate plants. (A. Whitney Brown) ________________________________________________________ Current beta is 3.61.09 (Echo) | 'Using TBBETA' information: http://www.silverstones.com/thebat/TBUDLInfo.html IMPORTANT: To register as a Beta tester, use this link first - http://www.ritlabs.com/en/partners/testers/
