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

Posted to dev@mahout.apache.org by Florents Tselai <fl...@gmail.com> on 2012/10/12 21:21:02 UTC

Clustering: Relative importance?

Hello everyone,

I'm using Mahout for clustering Market Basket data. I've previously used
SQL Server and have had some interesting results.
I would like to know if there is a way to calculate the "relative
importance" index when using mahout clustering algorithms.

Relative importance:  indicates how important the attribute and value pair
> is as a distinguishing factor. The longer the bar, the more likely it is
> that this attribute is representative of this category.


You may look here an example using this indicator:
http://technet.microsoft.com/en-us/library/dd299419%28v=sql.100%29.aspx