From craig@deersoft.com  Fri Aug 23 11:07:09 2002
Return-Path: <craig@deersoft.com>
Delivered-To: yyyy@localhost.netnoteinc.com
Received: from localhost (localhost [127.0.0.1])
	by phobos.labs.netnoteinc.com (Postfix) with ESMTP id 31B4044166
	for <jm@localhost>; Fri, 23 Aug 2002 06:05:15 -0400 (EDT)
Received: from phobos [127.0.0.1]
	by localhost with IMAP (fetchmail-5.9.0)
	for jm@localhost (single-drop); Fri, 23 Aug 2002 11:05:15 +0100 (IST)
Received: from mclean.mail.mindspring.net (mclean.mail.mindspring.net
    [207.69.200.57]) by dogma.slashnull.org (8.11.6/8.11.6) with ESMTP id
    g7N5rgZ12375 for <jm@jmason.org>; Fri, 23 Aug 2002 06:53:42 +0100
Received: from user-105nd99.dialup.mindspring.com ([64.91.181.41]
    helo=belphegore.hughes-family.org) by mclean.mail.mindspring.net with
    esmtp (Exim 3.33 #1) id 17i7OE-00023R-00; Fri, 23 Aug 2002 01:53:46 -0400
Received: from balam.hughes-family.org (balam.hughes-family.org
    [10.0.240.3]) by belphegore.hughes-family.org (Postfix) with ESMTP id
    63E8D71672; Thu, 22 Aug 2002 22:53:44 -0700 (PDT)
Date: Thu, 22 Aug 2002 22:53:44 -0700
Subject: Re: results for giant mass-check (phew)
Content-Type: text/plain; charset=US-ASCII; format=flowed
MIME-Version: 1.0 (Apple Message framework v482)
Cc: yyyy@example.com (Justin Mason),
	"Malte S. Stretz" <msquadrat.nospamplease@gmx.net>,
	SpamAssassin Talk ML <spamassassin-talk@lists.sourceforge.net>
To: Scott A Crosby <scrosby@cs.rice.edu>
From: "Craig R.Hughes" <craig@deersoft.com>
In-Reply-To: <oydit2248oi.fsf@bert.cs.rice.edu>
Message-Id: <AF40EEDE-B65C-11D6-8F61-00039396ECF2@deersoft.com>
Content-Transfer-Encoding: 7bit
X-Mailer: Apple Mail (2.482)
X-Pyzor: Reported 0 times.
X-Spam-Status: No, hits=-5.6 required=7.0
	tests=EMAIL_ATTRIBUTION,FORGED_RCVD_TRAIL,IN_REP_TO,
	      QUOTED_EMAIL_TEXT,SPAM_PHRASE_00_01,USER_AGENT_APPLEMAIL
	version=2.40-cvs
X-Spam-Level: 

I never claimed it could learn *all* combinatorial 
possibilities, but it certainly can learn some.

C

On Thursday, August 22, 2002, at 07:21  PM, Scott A Crosby wrote:

> On Thu, 22 Aug 2002 09:20:05 -0700, Craig R.Hughes 
> <craig@deersoft.com> writes:
>
>> It *can* learn combinatorial stuff in a more subtle way.  Imagine
>
> No it can't..
>
> It can learn a few examples that happen to be linearily seperable,
> like those you gave. It cannot learn the example I gave below, which
> is not linearily seperable.