From anonymoususr at gmail.com  Tue Mar  3 23:08:42 2009
From: anonymoususr at gmail.com (mbn)
Date: Tue, 3 Mar 2009 14:08:42 -0800 (PST)
Subject: [spambayes-dev]  Geometric Mean calculation
Message-ID: <22319020.post@talk.nabble.com>


Hi Guys,

Sorry for reviving an old topic, but I would like to get more information
about the geometric mean (GM) method devised by G. Robinson. The earliest
reference I could find is his article:

http://radio.weblogs.com/0101454/stories/2002/09/16/spamDetection.html

Where he mentions that the geometric mean is consistent with Fisher's
method, but no more details are given:

"The connection would take too much time to explain here, but it is a great
justification for why this should work really well."

I understant GM is now obsolete and has been replaced in favor of Fisher's,
but I am interested in the derivation of the method (or at least the basic
intuition behind it) and its relationship with Fisher's. I could not find
any explanation in the archives of SpamBayes. 

The most similar thing to Robinson's GM I could find is "logarithmic opinion
pool" with weights wi=1/n and constant k=1, but I have no idea whether the
two things are related. 

Any explanation about the derivation of the GM method will be greatly
appreciated.

Thanks in advance

P.S. I sent this same question to the [Spambayes - General] list and
somebody suggested me to write here.

-- 
View this message in context: http://www.nabble.com/Geometric-Mean-calculation-tp22319020p22319020.html
Sent from the Spambayes - Development mailing list archive at Nabble.com.


From skippy.hammond at gmail.com  Wed Mar  4 12:02:08 2009
From: skippy.hammond at gmail.com (Mark Hammond)
Date: Wed, 04 Mar 2009 22:02:08 +1100
Subject: [spambayes-dev] new spambayes build?
In-Reply-To: <18846.18862.344956.305457@montanaro.dyndns.org>
References: <499D2360.1070509@gmail.com>
	<18846.18862.344956.305457@montanaro.dyndns.org>
Message-ID: <49AE5FB0.8080500@gmail.com>

On 20/02/2009 5:11 PM, skip at pobox.com wrote:
>      Mark>  I recently found and fixed the problem in pywin32 which caused the
>      Mark>  DEP errors, and is probably responsible for many of the outlook
>      Mark>  related reports of crashing.  I'm happy to put a new windows
>      Mark>  binary out after all these years if anyone else is up for
>      Mark>  coordinating other releases.
>
>      Mark>  I'm aware it is almost a year since I discussed a new release with
>      Mark>  Skip at pycon ... :)
>
> PyCon is coming up.  Will you be there again?

I'm sure you have worked out the answer is 'no' by now ;)

> Given that I have reently checked in a fairly large set of changes I think
> I'd like it to cook a little bit more before making another release.

I'm happy with not making a 'real' release, but we should get a new 
binary out there for as much testing as possible so we *can* be 
confident about a real one.

I've a new windows binary from svn (so is branded however svn is - I 
can't recall) - what should I do with it?

Cheers,

Mark