Data Mining Techniques for Segmentation - Data Mining Techniques in CRM: Inside Customer Segmentation - page 108

Figure 3.19 Comparing the cluster's mean values to the overall mean values through
the SPSS AIM command.
standardized ( z -scores) in order to reduce the measurement scale differences and
more clearly portray the structure of the clusters.
All the above sets of charts, although not discussed in detail, reinforce and
further clarify the initial cluster interpretation. Based on the profiling information,
it is up to the project team involved to decide on the effectiveness of the solution.
If the solution makes sense and covers the business objectives it will be adopted
and deployed in the organization to support the marketing strategies. If not,
the segmentation process should be reviewed to search for a more effective
segmentation scheme.
SELECTING THE OPTIMAL CLUSTER SOLUTION
Regardless of the clustering method, data miners should also assess the homogene-
ity/cohesion of the clusters and the level of similarity of their members, as well as
their separation, by examining the measures presented in the relevant section.