org.kie.pmml.models.regression.model

Class KiePMMLRegressionClassificationTable

java.lang.Object

org.kie.pmml.models.regression.model.KiePMMLRegressionTable

org.kie.pmml.models.regression.model.KiePMMLRegressionClassificationTable

All Implemented Interfaces:

Serializable

public abstract class KiePMMLRegressionClassificationTable
extends KiePMMLRegressionTable

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field and Description
protected Map<String,KiePMMLRegressionTable>	categoryTableMap
protected OP_TYPE	opType
protected REGRESSION_NORMALIZATION_METHOD	regressionNormalizationMethod

Fields inherited from class org.kie.pmml.models.regression.model.KiePMMLRegressionTable

categoricalFunctionMap, intercept, numericFunctionMap, outputFieldsMap, predictorTermsFunctionMap, targetField

Constructor Summary

Constructors

Constructor and Description
KiePMMLRegressionClassificationTable()

Method Summary

Modifier and Type	Method and Description
Object	evaluateRegression(Map<String,Object> input)
Map<String,KiePMMLRegressionTable>	getCategoryTableMap()
OP_TYPE	getOpType()
protected abstract LinkedHashMap<String,Double>	getProbabilityMap(LinkedHashMap<String,Double> resultMap)
protected LinkedHashMap<String,Double>	getProbabilityMap(LinkedHashMap<String,Double> resultMap, DoubleUnaryOperator firstItemOperator, DoubleUnaryOperator secondItemOperator)
REGRESSION_NORMALIZATION_METHOD	getRegressionNormalizationMethod()
abstract boolean	isBinary() A Classification is considered binary if it is of CATEGORICAL type and contains exactly two Regression tables
protected abstract void	populateOutputFieldsMapWithProbability(Map.Entry<String,Double> predictedEntry, LinkedHashMap<String,Double> probabilityMap)
protected void	updateResult(AtomicReference<Double> toUpdate)

Methods inherited from class org.kie.pmml.models.regression.model.KiePMMLRegressionTable

getCategoricalFunctionMap, getIntercept, getNumericFunctionMap, getOutputFieldsMap, getPredictorTermsFunctionMap, getTargetCategory, getTargetField, populateOutputFieldsMapWithResult

Methods inherited from class java.lang.Object

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

Field Detail

regressionNormalizationMethod

protected REGRESSION_NORMALIZATION_METHOD regressionNormalizationMethod

opType

protected OP_TYPE opType

categoryTableMap

protected Map<String,KiePMMLRegressionTable> categoryTableMap

Constructor Detail

KiePMMLRegressionClassificationTable

public KiePMMLRegressionClassificationTable()

Method Detail

evaluateRegression

public Object evaluateRegression(Map<String,Object> input)

Overrides:

evaluateRegression in class KiePMMLRegressionTable

isBinary

public abstract boolean isBinary()

A Classification is considered binary if it is of CATEGORICAL type and contains exactly two Regression tables

Returns:

getProbabilityMap

protected abstract LinkedHashMap<String,Double> getProbabilityMap(LinkedHashMap<String,Double> resultMap)

populateOutputFieldsMapWithProbability

protected abstract void populateOutputFieldsMapWithProbability(Map.Entry<String,Double> predictedEntry,
                                                               LinkedHashMap<String,Double> probabilityMap)

updateResult

protected void updateResult(AtomicReference<Double> toUpdate)

Specified by:

updateResult in class KiePMMLRegressionTable

getRegressionNormalizationMethod

public REGRESSION_NORMALIZATION_METHOD getRegressionNormalizationMethod()

getOpType

public OP_TYPE getOpType()

getCategoryTableMap

public Map<String,KiePMMLRegressionTable> getCategoryTableMap()

getProbabilityMap

protected LinkedHashMap<String,Double> getProbabilityMap(LinkedHashMap<String,Double> resultMap,
                                                         DoubleUnaryOperator firstItemOperator,
                                                         DoubleUnaryOperator secondItemOperator)