From spamassassin-devel-admin@lists.sourceforge.net Sun Sep 22 00:47:04 2002 Return-Path: Delivered-To: yyyy@localhost.spamassassin.taint.org Received: from localhost (jalapeno [127.0.0.1]) by jmason.org (Postfix) with ESMTP id B748916F03 for ; Sun, 22 Sep 2002 00:47:03 +0100 (IST) Received: from jalapeno [127.0.0.1] by localhost with IMAP (fetchmail-5.9.0) for jm@localhost (single-drop); Sun, 22 Sep 2002 00:47:03 +0100 (IST) Received: from usw-sf-list2.sourceforge.net (usw-sf-fw2.sourceforge.net [216.136.171.252]) by dogma.slashnull.org (8.11.6/8.11.6) with ESMTP id g8LNgUC09546 for ; Sun, 22 Sep 2002 00:42:30 +0100 Received: from usw-sf-list1-b.sourceforge.net ([10.3.1.13] helo=usw-sf-list1.sourceforge.net) by usw-sf-list2.sourceforge.net with esmtp (Exim 3.31-VA-mm2 #1 (Debian)) id 17stsz-0000Zy-00; Sat, 21 Sep 2002 16:42:05 -0700 Received: from blount.mail.mindspring.net ([207.69.200.226]) by usw-sf-list1.sourceforge.net with esmtp (Exim 3.31-VA-mm2 #1 (Debian)) id 17stsm-0001nh-00 for ; Sat, 21 Sep 2002 16:41:52 -0700 Received: from user-vcaur02.dsl.mindspring.com ([216.175.108.2] helo=belphegore.hughes-family.org) by blount.mail.mindspring.net with esmtp (Exim 3.33 #1) id 17stsj-0000XM-00 for spamassassin-devel@lists.sourceforge.net; Sat, 21 Sep 2002 19:41:49 -0400 Received: by belphegore.hughes-family.org (Postfix, from userid 48) id 21E8791DC7; Sat, 21 Sep 2002 16:41:49 -0700 (PDT) From: bugzilla-daemon@hughes-family.org To: spamassassin-devel@example.sourceforge.net X-Bugzilla-Reason: AssignedTo Message-Id: <20020921234149.21E8791DC7@belphegore.hughes-family.org> Subject: [SAdev] [Bug 1012] negate directive addition Sender: spamassassin-devel-admin@example.sourceforge.net Errors-To: spamassassin-devel-admin@example.sourceforge.net X-Beenthere: spamassassin-devel@example.sourceforge.net X-Mailman-Version: 2.0.9-sf.net Precedence: bulk List-Help: List-Post: List-Subscribe: , List-Id: SpamAssassin Developers List-Unsubscribe: , List-Archive: X-Original-Date: Sat, 21 Sep 2002 16:41:49 -0700 (PDT) Date: Sat, 21 Sep 2002 16:41:49 -0700 (PDT) http://www.hughes-family.org/bugzilla/show_bug.cgi?id=1012 ------- Additional Comments From easmith@beatrice.rutgers.edu 2002-09-21 16:41 ------- I like this idea, but from my reading - part of my dissertation research - on GAs (or, in this case, more precisely "Evolutionary Programming", since the parameters are non-binary and the major operator is mutation, not crossover), you do need to keep in mind that the more interactions there are between different variables, the better-tweaked the GA/EA will need to be, especially to avoid local optima (which may have been the problem with the suspiciously-high "anti-ratware" USER_AGENT scores). From past results, what's needed are one or more of the below: A. Determine mutational parameters by adapted-scores themselves, with variation on a per-original score basis; ideally, allow for having "correlated" mutations - in other words, have a mechanism in place for trying out changes to a bunch of scores at once, with them all moving about the same amount (albeit possibly in different directions) - one means to do this is found in the "Evolutionary Strategies" of Schwefel; B. Adapt the probability of a mutation taking place depending on how well previous mutation tries have done - if more than about a fifth of the new "individuals" are doing about as well as, or better than, the parental generation, then mutate more parameters; if less than a fifth of them are doing about as well as, or better than, the parental generation, then mutate less parameters. -Allen ------- You are receiving this mail because: ------- You are the assignee for the bug, or are watching the assignee. ------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf _______________________________________________ Spamassassin-devel mailing list Spamassassin-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/spamassassin-devel