On Fri, Nov 14, 2003 at 10:25:54AM -0500, Matt Kettler wrote:
> Yes. SA won't even try to use bayes scoring until 200 spam AND 200 ham 
> messages are trained.

... not to mention that only learning spam means bayes will be skewed
to think everything is spam.

-- 
Randomly Generated Tagline:
Uh, so.  Let's have a conversation.  Uh, I think we'll find that we have
 very little in common.
 
                -- Homer Simpson
                   The Last Temptation of Homer

Attachment: pgp00000.pgp
Description: PGP signature

Reply via email to