CLUSTER_PROBABILITY

Syntax

Description of the illustration cluster_probability.gif

mining_attribute_clause::=

Description of the illustration mining_attribute_clause.gif

Purpose

This function is for use with clustering models that have been created with the DBMS_DATA_MINING package or with the Oracle Data Mining Java API. It returns a measure of the degree of confidence of membership of an input row in a cluster associated with the specified model.

• For cluster_id, specify the identifier of the cluster in the model. The function returns the probability for the specified cluster. If you omit this clause, then the function returns the probability associated with the best predicted cluster. You can use the form without cluster_id in conjunction with the CLUSTER_ID function to obtain the best predicted pair of cluster ID and probability.

• The mining_attribute_clause behaves as described for the PREDICTION function. Refer to mining_attribute_clause

Examples

The following example determines the ten most representative customers, based on likelihood, in cluster 2.

This example, and the prerequisite data mining operations, including the creation of the dm_sh_clus_sample model and the dm_sh_sample_apply_prepared view, can be found in the demo file $ORACLE_HOME/rdbms/demo/dmkmdemo.sql. General information on data mining demo files is available in Oracle Data Mining Administrator's Guide. The example is presented here to illustrate the syntactic use of the function.

SELECT *
  FROM (SELECT cust_id, CLUSTER_PROBABILITY(km_sh_clus_sample, 2 USING *) prob
          FROM km_sh_sample_apply_prepared
        ORDER BY prob DESC)
 WHERE ROWNUM < 11;

   CUST_ID   PROB
---------- ------
    100052  .9993
    100962  .9993
    101208  .9993
    100281  .9993
    100012  .9993
    101009  .9992
    100173  .9992
    101176  .9991
    100672  .9991
    101420  .9991
 
10 rows selected.