R help - May 2005 - standardization

SAS Enterprise Miner recommendeds to standardize using X / STDEV(X) 
versus [X ? mean(X)] / STDEV(X)

Any thoughts on this? Pros Cons

Philip

Philip Bermingham <pberming at research.ryerson.ca> writes:
> SAS Enterprise Miner recommendeds to standardize using X / STDEV(X)
> versus [X ? mean(X)] / STDEV(X)
> 
> Any thoughts on this? Pros Cons
When??? 

This makes absolutely no sense out of context.

-- 
   O__  ---- Peter Dalgaard             Blegdamsvej 3  
  c/ /'_ --- Dept. of Biostatistics     2200 Cph. N   
 (*) \(*) -- University of Copenhagen   Denmark      Ph: (+45) 35327918
~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk)             FAX: (+45) 35327907

My thoughts on this is:
Do not trust what SAS say??s and least of all what the Enterprise Miner said.

Robust Statisticians recommendends to standardize using e.g. 
(X - median(X)) / ( MAD(X) / 0.675 ) 

Best,
Matthias

> SAS Enterprise Miner recommendeds to standardize using X / STDEV(X) 
> versus [X - mean(X)] / STDEV(X)
> 
> Any thoughts on this? Pros Cons
> 
> Philip
> 
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list 
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read 
> the posting guide! http://www.R-project.org/posting-guide.html
>

> Robust Statisticians recommendends to standardize using e.g. 
> (X - median(X)) / ( MAD(X) / 0.675 ) 
Make that (X - median(X))/mad(X)

The constant is already in mad()


-- Bert Gunter
Genentech Non-Clinical Statistics
South San Francisco, CA
 
"The business of the statistician is to catalyze the scientific learning
process."  - George E. P. Box

You asked another question about clustering, so I presume you want to
standardize some variables before clustering. In SAS, PROC STDIZE
offers 18 standardization methods. See
http://support.sas.com/91doc/getDoc/statug.hlp/stdize_sect12.htm#stat_stdize_stdizesm
for details. If you're really concerned about this I would suggest
running simulations to compare the performance of various
standardization methods (relative to your data and what you're after).

hth,
b.


-----Original Message-----
From: Philip Bermingham [mailto:pberming at research.ryerson.ca]
Sent: Wednesday, May 18, 2005 8:34 AM
To: r-help at stat.math.ethz.ch
Subject: [R] standardization


SAS Enterprise Miner recommendeds to standardize using X / STDEV(X) 
versus [X ? mean(X)] / STDEV(X)

Any thoughts on this? Pros Cons

Philip

______________________________________________
R-help at stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html