Matt Kettler <[EMAIL PROTECTED]> writes:
> At 07:31 AM 9/10/2004, Gustafson, Tim wrote:
> >What I'm worried about is that I have so many more SPAM than HAM messages.
> >Is this dangerous?
>
> No, in fact it's closer to optimal than a 50-50 mix is...
>
> Remember, Bayes is a statistical system..
At 07:31 AM 9/10/2004, Gustafson, Tim wrote:
What I'm worried about is that I have so many more SPAM than HAM messages.
Is this dangerous?
No, in fact it's closer to optimal than a 50-50 mix is...
Remember, Bayes is a statistical system.. Statistics work best when they
are as close to reality as p
Are you autotraining or manually training? If the former try manual
training. I've never used the automatic training here. It seems to be
based on circular logic. It reinforces initial bad guesses about what
is spam and what is ham. And right off on an install has Spam Assassin
er ah "not doing ver
Hello
My Bayes filter has been learning beautifully. There is just one problem:
more than 70% of my e-mail, on average, gets tagged as SPAM. In just 13
days of Bayes auto-learning, I have amassed the following SPAM/HAM messages
(via the sa-learn --dump magic command):
0.000 0