Clustering Apply Option

Specify the output option for clustering apply.

The rules displayed in the model build activity do not necessarily assign each point in the data space to a unique cluster. For example, the model will associate a new customer with an existing cluster, even if the new customer doesn't conform exactly to the rules for any cluster. Thus a probability of membership in each cluster is assigned to a record. The default is for the apply output to show only the cluster with highest probability of membership for a given record. Otherwise, you display the probabilities for any number of clusters, either by specifying particular clusters, or by indicating the number of clusters ranked by probability.

The wizard displays the number of leaf clusters.

You can choose one of the following options:

Click Next to continue.