CLASS-L Archives

December 2004

CLASS-L@LISTS.SUNYSB.EDU

Options: Use Monospaced Font
Show Text Part by Default
Condense Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Mime-Version:
1.0
Sender:
"Classification, clustering, and phylogeny estimation" <[log in to unmask]>
Subject:
From:
Alexander Sirotkin <[log in to unmask]>
Date:
Fri, 24 Dec 2004 13:13:01 +0200
Content-Transfer-Encoding:
7bit
Content-Type:
text/plain; charset=US-ASCII
Reply-To:
"Classification, clustering, and phylogeny estimation" <[log in to unmask]>
Parts/Attachments:
text/plain (20 lines)
Hi.

In my master thesis I have to compare two clusterings. There are a
number of measures one can use to do this, the most common probably
being Rand index. The major problem I have with Rand index is that I
don't know what is the critical value I should compare it to. Some
people suggest 0.7 as a rule of thumb without any theoretical
explanation.

I found a paper
http://www.quantlet.de/scripts/compstat2002_wh/paper/full/C_06_saporta.pdf
which claims that Rand index is asymptotically normal and they have a
formula for variance, which produces weird results in my case (sample
of size 2000, 75 classes).

Can somebody give me some tips on the subject ?

--
Alexander Sirotkin

ATOM RSS1 RSS2