I don't know if this was discussed. K-nearest neighbors regression
and neural networks are prediction procedures for which the
inputs should be scaled. There are probably others. Whatever
software you are running, you need to check the documentation
to see whether input transformations are handled for you within
the procedure or whether you have to apply transformations yourself.
Tony Babinec
-----Original Message-----
From: SPSSX(r) Discussion [mailto:
[hidden email]] On Behalf Of Chi
Shu
Sent: Monday, October 14, 2013 7:27 PM
To:
[hidden email]
Subject: The reason to perform linear transformation of independent
variables
Dear Listers,
In ADP node, there is an option to transform continuous independent
variables using either z-score or min/max method.
I can see the point of transforming dependent variable for conforming
correct the distribution of error terms, but what's the reason to transform
the variables on the right side of the equal sign, in other words, what's
advantage of rescaling IV using linear transformation?
Thanks!
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command SIGNOFF SPSSX-L For a list of
commands to manage subscriptions, send the command INFO REFCARD
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
Locked
