public class NeuralNetworkEvaluator extends NeuralNetworkManager implements Evaluator
| Constructor and Description |
|---|
NeuralNetworkEvaluator(PMML pmml) |
NeuralNetworkEvaluator(PMML pmml,
NeuralNetwork neuralNetwork) |
| Modifier and Type | Method and Description |
|---|---|
Map<FieldName,?> |
evaluate(Map<FieldName,?> arguments) |
Map<String,Double> |
evaluateRaw(EvaluationContext context)
Evaluates neural network.
|
com.google.common.collect.BiMap<String,Entity> |
getEntityRegistry() |
Object |
prepare(FieldName name,
Object value)
Prepares the input value for a field.
|
addConnection, addConnection, addNeuralInput, addNeuralLayer, addNeuralOutput, addNeuron, createModel, getModel, getNeuralInputs, getNeuralLayers, getOrCreateNeuralOutputs, getSummaryaddField, addMiningField, getActiveFields, getGroupFields, getMiningField, getMiningFields, getMiningSchema, getOrCreateLocalTransformations, getOrCreateOutput, getOrCreateTargets, getOutputField, getOutputFields, getPredictedFields, getTarget, getTargetField, resolveFieldaddDataField, find, find, findAll, getDataDictionary, getDataField, getHeader, getModel, getModelManager, getModelManager, getModels, getOrCreateTransformationDictionary, getPmml, resolveFunctionclone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, waitgetActiveFields, getDataField, getGroupFields, getMiningField, getOutputField, getOutputFields, getPredictedFields, getSummary, getTargetFieldpublic NeuralNetworkEvaluator(PMML pmml)
public NeuralNetworkEvaluator(PMML pmml, NeuralNetwork neuralNetwork)
public com.google.common.collect.BiMap<String,Entity> getEntityRegistry()
getEntityRegistry in interface HasEntityRegistry<Entity>getEntityRegistry in class NeuralNetworkManagerpublic Object prepare(FieldName name, Object value)
Evaluatorprepare in interface Evaluatorname - The name of the fieldConsumer.getDataField(FieldName),
Consumer.getMiningField(FieldName)public Map<FieldName,?> evaluate(Map<FieldName,?> arguments)
evaluate in interface Evaluatorarguments - Map of active field values.predicted field and output field values.
Simple values are represented using the Java equivalents of PMML data types (eg. String, Integer, Float, Double etc.).
Complex values are represented as instances of Computable that return simple values.Computablepublic Map<String,Double> evaluateRaw(EvaluationContext context)
NeuralInput.getId(),
Neuron.getId()Copyright © 2013 University of Tartu. All Rights Reserved.