You are viewing a plain text version of this content. The canonical link for it is here.

Posted to user@mahout.apache.org by sam wu <sw...@gmail.com> on 2013/01/28 04:49:02 UTC

normalize feature, scatterplot correlation

Hi,
After normalizing some features, I did some clustering.
After getting clustering label, I try to examine the correlation between
features using R scatterPlot conditional on the cluster label.

The question is
should I use the original un-normalized data or post-normalized data for
comparing scatterplot correlation ?

This seems not to be directed related with Mahout, but idea is highly
appreciated.

BR

Sam