ClusterNaming

When a clustering data mining model is processed, the resulting clusters are named Cluster 1, Cluster 2, etc. The ClusterNaming functions allow you to quickly determine the characteristics that distinguish each cluster from the whole population.

In the Customer Clusters mining model in Adventure Works, the distinguishing characteristics of Cluster 1 are Home Owner = Yes; Commute Distance = 0-1 Miles; Education = Graduate Degree. However, the distinguishing characteristics for Cluster 3 are Marital Status = Single; Yearly Income = 10000 - 30000; Home Owner = No.

The following DMX query, when run against the Adventure Works DW database, returns a recordset describing as succinctly as possible the distinguishing characteristics of each cluster:

CALL ASSP.DistinguishingCharacteristicsForClusters("Customer Clusters")

The following DMX query, when run against the Adventure Works DW database, renames each cluster so the name reflects that cluster's distinguishing characteristics:

CALL ASSP.AutoNameClusters("Customer Clusters")


General Note: The ASSP assembly must be registered with unrestricted permissions for the ClusterNaming functions to succeed.

Last edited Jun 22, 2007 at 11:51 PM by dgosbell, version 4