PREDICTION_BOUNDS

Syntax

Description of the illustration ''prediction_bounds.gif''

mining_attribute_clause::=

Description of the illustration ''mining_attribute_clause.gif''

Purpose

The PREDICTION_BOUNDS function is for use with generalized linear models (GLM) created by the DBMS_DATA_MINING package or with Oracle Data Miner. It returns an object with two NUMBER fields LOWER and UPPER. For a regression mining function, the bounds apply to value of the prediction. For a classification mining function, the bounds apply to the probability value. If the GLM was built using ridge regression, or if the covariance matrix is found to be singular during the build, then this function returns NULL for both fields.

• For confidence_level, specify a number in the range (0,1). If you omit this clause, then the default value is 0.95.

• The class_value argument is valid for classification models but not for regression models. By default, the function returns the bounds for the prediction with the highest probability. You can use the class_value argument to filter out the bounds value specific to a target value.

  You can specify class_value while leaving confidence_level at its default by specifying NULL for confidence_level.

• The mining_attribute_clause has the same behavior for PREDICTION_BOUNDS that it has for PREDICTION. Refer to mining_attribute_clause.

Example

The following example returns the distribution of customers whose ages are predicted to be between 25 and 45 years with 98% confidence.

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

SELECT count(cust_id) cust_count, cust_marital_status
  FROM (SELECT cust_id, cust_marital_status
    FROM mining_data_apply_v
    WHERE PREDICTION_BOUNDS(glmr_sh_regr_sample,0.98 USING *).LOWER > 24 AND
          PREDICTION_BOUNDS(glmr_sh_regr_sample,0.98 USING *).UPPER < 46)
    GROUP BY cust_marital_status;
 
    CUST_COUNT CUST_MARITAL_STATUS
-------------- --------------------
            46 NeverM
             7 Mabsent
             5 Separ.
            35 Divorc.
            72 Married