Oracle Data Mining Java API Reference
10g Release 1 (10.1)

B12276-01

oracle.dmt.odm.task
Class CrossValidateTask

java.lang.Object
  |
  +--oracle.dmt.odm.MiningObject
        |
        +--oracle.dmt.odm.LocatableObject
              |
              +--oracle.dmt.odm.task.MiningTask
                    |
                    +--oracle.dmt.odm.task.MiningDataTask
                          |
                          +--oracle.dmt.odm.task.ClassificationTestTask
                                |
                                +--oracle.dmt.odm.task.CrossValidateTask
All Implemented Interfaces:
java.io.Serializable

public class CrossValidateTask
extends ClassificationTestTask

An instance of CrossValidateTask provides an additional technique for measuring the accuracy of a predictive model. Leave-one-out cross-validation is an especially accurate case of cross-validation, but it is ordinarily computationally expensive. In the case of Naive Bayes models, however, there is an efficient technique for leave-one-out cross-validation.

CrossValidateTask's input data must have all the active attributes specified in the associated cross-validate function settings. If the task input data is not compatible with the specified input model, then execution of CrossValidateTask fails with an exception.

After constructing a valid CrossValidateTask instance, it must be persisted in the database using store instance method before executing. Task name must be unique in the user schema task name space.

After successful completion of the cross-validate task, the user can retrieve the results using the ClassificationTestResult object. To remove the cross-validate task object in a database, call the static method remove(dmsConn, taskName).

Since:
ODM 9.2.0
See Also:
Serialized Form

Constructor Summary
CrossValidateTask(PhysicalDataSpecification taskInputData, java.lang.String crossValidateSettingsName, java.lang.String crossValidateResultName)
          Creates an instance of test task from the given test data, the model to which the test data is applied, and the location of the test output.

 

Method Summary
 java.lang.String getSettingsName()
          Returns the name of the cross-validate settings name.

 

Methods inherited from class oracle.dmt.odm.task.ClassificationTestTask
getTestResultName, restore

 

Methods inherited from class oracle.dmt.odm.task.MiningDataTask
getInputModelName, getTaskInputData

 

Methods inherited from class oracle.dmt.odm.task.MiningTask
execute, getActiveExecutionHandle, getCurrentStatus, getExecutionDuration, getExecutionHandle, getLatestExecutionHandle, getName, getStatusHistory, getTaskType, getType, listTasks, remove, rename, store, terminate, waitForCompletion, waitForCompletion

 

Methods inherited from class java.lang.Object
equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

 

Constructor Detail

CrossValidateTask

public CrossValidateTask(PhysicalDataSpecification taskInputData,
                         java.lang.String crossValidateSettingsName,
                         java.lang.String crossValidateResultName)
                  throws InvalidArgumentException,
ODMException
Creates an instance of test task from the given test data, the model to which the test data is applied, and the location of the test output.
Parameters:
taskInputData - Physical data specification of the test input data
crossValidateSettingsName - Classification function settings to perform cross-validate
crossValidateResultName - Name of the cross-validate results
Throws:
InvalidArgumentException - is thrown
- when any of the parameters are null
- when the taskInputData does not have location access details
- when the crossValidateSettingsName/testResultName length >64
InvalidArgumentException
ODMException
Method Detail

getSettingsName

public java.lang.String getSettingsName()
Returns the name of the cross-validate settings name.

Copyright © 2003 Oracle Corporation. All Rights Reserved.