Skip navigation links

oracle.dmt.jdm.supervised.classification
Class OraClassificationSettings

java.lang.Object
oracle.dmt.jdm.OraDMObject
oracle.dmt.jdm.OraMiningObject
oracle.dmt.jdm.base.OraBuildSettings
oracle.dmt.jdm.supervised.OraSupervisedSettings
oracle.dmt.jdm.supervised.classification.OraClassificationSettings

All Implemented Interfaces:

javax.datamining.base.BuildSettings, javax.datamining.supervised.classification.ClassificationSettings, javax.datamining.MiningObject, oracle.dmt.jdm.OraPLSQLConstants, javax.datamining.supervised.SupervisedSettings

public class OraClassificationSettings

extends oracle.dmt.jdm.supervised.OraSupervisedSettings

implements javax.datamining.supervised.classification.ClassificationSettings

A ClassificationSettings instance supports function settings specific to the classification mining function.
For one-class classification models, the algorithm must be set to SVM and the target must be set to null. One-class classification implements the anomaly detection function, which identifies unusual patterns in the data. When used for anomaly detection, the SVM algorithm is known as One-Class SVM.

See Also:

MiningObject, oracle.dmt.jdm.MiningObjectImpl, BuildSettings, oracle.dmt.jdm.base.BuildSettingsImpl, SupervisedSettings, OraSupervisedSettings, ClassificationSettings

Method Summary

java.lang.String	getPriorProbabilitiesName() Returns the table name of the prior or null if the default prior table name is to be used.
java.util.Map	getTargetWeightsMap() Returns the map that contains the weights associated with the target classes.
java.lang.String	getTargetWeightsTableName() Returns the name of the table that contains the weights associated with the target classes.
void	setPriorProbabilitiesName(java.lang.String priorTableName) Sets the table name of a prior that will be used for model build.
void	setTargetWeightsMap(java.util.Map targetWeightsMap) Sets the map that contains the weights associated with the target classes.
void	setTargetWeightsTableName(java.lang.String targetWeightsTable) Sets the name of the table that contains the weights associated with the target classes.

Methods inherited from class oracle.dmt.jdm.supervised.OraSupervisedSettings

getTargetAttributeName, setTargetAttributeName

Methods inherited from class oracle.dmt.jdm.base.OraBuildSettings

useAutomatedDataPreparations, useAutomatedDataPreparations

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Methods inherited from interface javax.datamining.supervised.SupervisedSettings

getTargetAttributeName, setTargetAttributeName

Methods inherited from interface javax.datamining.base.BuildSettings

getAlgorithmSettings, getAttributeNames, getLogicalAttributes, getLogicalData, getLogicalDataName, getMiningFunction, getOutlierIdentification, getOutlierTreatment, getUsage, getWeight, getWeightAttribute, setAlgorithmSettings, setLogicalDataName, setOutlierIdentification, setOutlierTreatment, setUsage, setWeight, setWeightAttribute, verify

Methods inherited from interface javax.datamining.MiningObject

getCreationDate, getCreatorInfo, getDescription, getName, getObjectIdentifier, getObjectType, setDescription

Method Detail

setPriorProbabilitiesName

public void setPriorProbabilitiesName(java.lang.String priorTableName)

Sets the table name of a prior that will be used for model build.

Parameters:

priorTableName - The table name of a prior to be associated with classification settings.

Returns:

void

getPriorProbabilitiesName

public java.lang.String getPriorProbabilitiesName()

Returns the table name of the prior or null if the default prior table name is to be used.

Returns:

String

setTargetWeightsTableName

public void setTargetWeightsTableName(java.lang.String targetWeightsTable)

Sets the name of the table that contains the weights associated with the target classes. By default all target classes are given equal weight. This table can be used to override the default weights. This is useful when some target classes are more important to the user than others. Thus the model can achieve higher accuracy on higher weighted classes. In 11g release only GLM algorithm supports this feature. Other classification algorithms can achieve the same feature using the Cost Matrix either in the model build or apply.

Parameters:

targetWeightsTable - Name of the table that contains the weights associated with the target classes.

Returns:

void

Since:

Oracle 11.0.0.0

getTargetWeightsTableName

public java.lang.String getTargetWeightsTableName()

Returns the name of the table that contains the weights associated with the target classes. In 11g release only GLM algorithms supports this feature.

Returns:

String

Since:

Oracle 11.0.0.0

setTargetWeightsMap

public void setTargetWeightsMap(java.util.Map targetWeightsMap)

Sets the map that contains the weights associated with the target classes. By default all target classes are given equal weight. This map can be used to override the default weights. This is useful when some target classes are more important to the user than others. Thus the model can achieve higher accuracy on higher weighted classes. In 11g release only GLM algorithm supports this feature. Other classification algorithms can achieve the same feature using the Cost Matrix either in the model build or apply.

Parameters:

targetWeightsMap -

Since:

Oracle 11.0.0.0

getTargetWeightsMap

public java.util.Map getTargetWeightsMap()

Returns the map that contains the weights associated with the target classes.

Since:

Oracle 11.0.0.0

Skip navigation links