comp.lang.idl-pvwave archive: archive

Home » Public Forums » archive » clustering

Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend

Re: clustering [message #54960 is a reply to message #54850]

Mon, 23 July 2007 11:16

James Kuyper
Messages: 425
Registered: March 2000

Senior Member

little davey wrote:
> Is it the case that you MUST use standardize() before you call
> CLUST_WTS()?

No, you don't need to standardize. The scaling of the data affects the
results you're going to get. The algorithm is built to treat a
difference of 1.0 in a variable between two data points as being
equally significant, no matter which of the variables that difference
occurs in. This is a fine assumption for variables that have
equivalent meanings; such as the x, y, and z coordinates when you're
clustering stars.

However, in most contexts for most variables that's simply not true.
You can use the scaling to tell CLUST_WTS() treat differences in one
variable as more important than differences in another variable.
That's fine if you have a clear idea as to which variables are more
important than others, and by what factor.

However, the most common case is where the analyst doesn't have clear
advance knowledge of the relative importance of the different
variables; the analysis is being done to get some idea as to which are
the important variables. Scaling the variables according their
standard deviations, as STANDARDIZE() does,provides a comforting
illusion of objectivity to the choice of scale factors. Unfortunately,
if there are only very small variations in an unimportant variable,
standardizing it would give that variable undue importance in the
clustering. Nothing can substitute for good judgment on the part of
the analyst. However, STANDARDIZE() does produce fairly good results
in many cases, which implies that the objectivity it provides not
quite as illusionary as I've suggested.

Report message to a moderator

[Message index]

		clustering By: nivedita.raghunath on Tue, 17 July 2007 08:36
		Re: clustering By: Vince Hradil on Thu, 19 July 2007 10:30
		Re: clustering By: nivedita.raghunath on Thu, 19 July 2007 09:12
		Re: clustering By: Vince Hradil on Thu, 19 July 2007 13:46
		Re: clustering By: Conor on Wed, 18 July 2007 12:07
		Re: clustering By: Vince Hradil on Mon, 23 July 2007 11:22
		Re: clustering By: James Kuyper on Mon, 23 July 2007 11:16
		Re: clustering By: little davey on Mon, 23 July 2007 10:14
		Re: clustering By: little davey on Mon, 23 July 2007 09:08
		Re: clustering By: Conor on Mon, 23 July 2007 06:49
		Re: clustering By: nivedita.raghunath on Mon, 23 July 2007 06:45
		Re: Clustering By: Kai Mühlbauer on Tue, 01 November 2011 03:01
		Re: Clustering By: Kai Muehlbauer on Mon, 31 October 2011 08:40
		Re: Clustering By: Jeremy Bailin on Tue, 01 November 2011 12:16
		Re: Clustering By: kidpix on Mon, 31 October 2011 04:56
		Re: Clustering By: Kai Mühlbauer on Fri, 28 October 2011 10:07
		Re: clustering By: kisCA on Wed, 07 March 2012 13:37
		Re: clustering By: Russell[1] on Wed, 07 March 2012 09:32

Previous Topic:	Re: Move data to structure fields without copying
Next Topic:	Annotate Uncertainty [Upper & Lower] on a plot?

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Wed Oct 08 19:34:14 PDT 2025

Total time taken to generate the page: 0.00440 seconds