Bug 3118 - bayes: error in our calculations
Summary: bayes: error in our calculations
Status: RESOLVED FIXED
Alias: None
Product: Spamassassin
Classification: Unclassified
Component: Learner (show other bugs)
Version: SVN Trunk (Latest Devel Version)
Hardware: Other other
: P1 major
Target Milestone: 3.0.0
Assignee: SpamAssassin Developer Mailing List
URL:
Whiteboard:
Keywords:
Depends on:
Blocks: 3208
  Show dependency tree
 
Reported: 2004-03-01 15:07 UTC by Justin Mason
Modified: 2004-03-23 10:58 UTC (History)
0 users


Attachment Type Modified Status Actions Submitter/CLA Status
Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Justin Mason 2004-03-01 15:07:35 UTC 
Henry reports:

(15:03:50) HenryCStern: but there is effectively an off-by-one error in your
naive Bayes calculations
(15:04:09) HenryCStern: if you derive everything using Bayes' theorem, you'll
see what I mean
(15:05:47) justinmason23: so we should multiply both $H and $S by ($nham /
$totalmsgcount) and ($nspam / $totalmsgcount) respectively to get correct figures?
(15:05:57) HenryCStern: yeah

he brought it up before, but we all seem to have forgotten it.  so let's file a
bug this time! ;)
Comment 1 Justin Mason 2004-03-11 22:03:43 UTC 
testing this now.
Comment 2 Justin Mason 2004-03-13 18:07:55 UTC 
see bug 2129 for results; basically, a win, so it's in.