On 03/01, Jason Bertoch wrote:
> I haven't spent any time over the years reading ruleqa results, but
> is it normal for a rule's S/O to bounce all over the place from week
> to week, or is that an indication of a problem in the testing

RuleQA results are incredibly noisy:

http://www.chaosreigns.com/dnswl/
(I recommend firefox due to inconsistent implementations of svg rendering.)

The last two graphs on that page are most relevant - spam and ham counts,
and percentage of corpora which is ham.  The data is only the weekly -net
runs.

I've been meaning to look into this to see if it correlates to some people
submitting their data inconsistently or something.

-- 
"I don't want people who want to dance, I want people who have to dance."
--George Balanchine
http://www.ChaosReigns.com

Reply via email to