FocalPoint Clustering Features 

About Clustan
Cluster Analysis
User Support
What's New
White Papers
Contact Us

There are eight unique features that differentiate FocalPoint Clustering from other k-means analysis programs:

      1.  Tests case order sensitivity in trials
      2.  Exact relocation test assures convergence
      3.  Saves top solutions ordered by goodness-of-fit
      4.  Choice of six starting strategies
      5.  Estimates reproducibility of each solution
      6.  Removes outliers and intermediates
      7.  Clusters around means or exemplars
      8.  Automatically re-weights the variables

1.  Clustering procedures are notoriously sensitive to the case order, and k-means analysis is no exception.  Whereas other programs produce only one final cluster solution, FocalPoint tests the sensitivity of different case orders and compiles a range of cluster solutions that can then be compared and evaluated.

Other clustering programs do not test case order sensitivity

2.  Because FocalPoint uses an exact relocation test on the Euclidean Sum of Squares, convergence is assured and goodness-of-fit is calculated.

    Other clustering programs do not necessarily converge, report the sum of squares, or measure how well the cluster solutions fit the data

3.  FocalPoint finds several "top solutions" and lets you decide which of these to progress for action as a cluster model.

    Other clustering programs do not list "top solutions" or give a choice of cluster model for further analysis

4.  There are six ways of specifying the start of a FocalPoint cluster analysis.  You can section a tree, extract cliques or select exemplars from it; choose seed points from another cluster solution; randomly assign clusters; or specify a segmentation.

Other clustering programs do not offer six starting options

5.  FocalPoint calculates the reproducibility of each cluster solution from its frequency of occurrence over a large number of random trials.

Other clustering programs do not estimate the reproducibility of each cluster solution found

6.  Sub-optimal, solutions often exhibit small or singleton clusters.  These "outliers" are remote from the densest area of data.  FocalPoint allows such outliers to be identified and removed from the cluster solution, so that they do not distort the cluster means.

Other clustering programs do not find outliers and intermediate cases and remove them automatically from the cluster solution

7.  Conventional k-means analysis forms clusters around means.  You can do this with FocalPoint or, alternatively, cluster around "exemplars" - the most typical cases in each cluster.  The advantage of clustering around exemplars is that the cluster centres are actual cases.

Other clustering programs do not cluster around actual cases

8.  When you have a cluster solution that is actionable, FocalPoint provides guidance on the important variables that discriminate between the clusters.  You can use this to revise the variable weights, thereby calibrating your cluster model to your key variables and reducing the effect of noise variables.

Other clustering programs do not revise variable weights

Full and trial versions of FocalPoint are available.  To find out more, ORDER ClustanGraphics on-line now.