K-Means Clustering

K-MEANS CLUSTERING Excel Add-In


What is it?

This Excel Add-In allows you to run a cluster analysis using the K-Means algorithm. Suppose you have a sample of no more than 20,000 consumers,described by no more than 100 continuous variables. K-means will group the consumers into clusters that are homogeneous in terms of the descriptor variables.

Is this for me?

Do you have a PCrunning Microsoft Office 2003 or older?
Do you have a solid working knowledge of Excel?
Have you ever taken a Marketing Research class?

If you answered yes to all three questions, this Add-in mightbe of interest to you.

How does it work?

This add-in implements the K-means clustering algorithm which you can find discussed in any Marketing Research textbook


How do I install it?Before you install this Add-in, make sure that you first install the Excel “Analysis Toolpak” and “Analysis Toolpak VBA” Add-ins, which come with the Office installation disk.

1. If you are readingthis, you must have already unzipped the K-means Clustering.ZIP file into a directory in your hard disk (for example, C:\K-means) . This directory should contain the following files
README.DOC - thisfile
KMEANSCLUSTER 2007.XLA - Excel Add-In file
KMEANSCLUSTER.XLA - Excel Add-In file
KMEANSDLL.DLL - dynamic link library used by the Add-In
EXAMPLE.XLS - a sample spreadsheet


2.Move the KMEANSDLL.DLL file to the Windows/System32 directory (or Windows/System if you are using Windows 7).

3.a If you are using Office 2007, you should use Excel Options/ Addins/ Manage Addinsand point it to where the KmeansCluster 2007.xlam file is located.


3b. If you are using Office 2003 or older, follow these directions. First, open the Excel spreadsheet containing the data to beanalyzed (or the example, if you wish), which should be at the same directory mentioned in step 1.



4. Select Tools in the menu, and drag the Add-Ins option.




Then, click on Browse,...