Re: Mixed 64-bit system GerH binaries / BillYscripts

   From: "Trever L. Adams" <[email protected]>

   Bill Yerazunis wrote:
   > Now, here's the downside: to get a good library interface, it might
   > be necessary to make _all_ classifiers require two-sided training.
   >
   >   
   Is this really wise? Won't this make it very slow.

   I am not against it ALLOWING, but REQUIRING?

Not at all.  In the good-versus-spam scenario, where you only want to
say "spam" or "not spam" (and thus have only one statistics file), it
will take precisely as long as the current system.  There's no loss at
all.  In some systems, there might even be a gain.

For N-way splits, you are right, but in return, each class's "in"
classifier algorithm can be tailored to be optimal for that class,
(it could even be a regex or a Bloom filter).  That's something
we can't do right now, in any way whatsoever.

      - Bill Yerazunis

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/