org.neuroph.eval

Class Evaluation

java.lang.Object

org.neuroph.eval.Evaluation

public final class Evaluation
extends Object

Evaluation service used to run different evaluators on trained neural network

Constructor Summary

Constructors

Constructor and Description
Evaluation()

Method Summary

Modifier and Type	Method and Description
void	addEvaluator(Evaluator evaluator)
EvaluationResult	evaluate(NeuralNetwork neuralNetwork, DataSet testSet) Runs evaluation procedure for given neural network and data set through all evaluatoors Evaluation results are stored in evaluators
<T extends Evaluator> T	getEvaluator(Class<T> type)
Map<Class<?>,Evaluator>	getEvaluators() Return all evaluators used for evaluation
double	getMeanSquareError()
static void	runFullEvaluation(NeuralNetwork<?> neuralNet, DataSet dataSet) Out of the box method (util) which computes all metrics for given neural network and test data set

Methods inherited from class java.lang.Object

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

Constructor Detail

Evaluation

public Evaluation()

Method Detail

evaluate

public EvaluationResult evaluate(NeuralNetwork neuralNetwork,
                                 DataSet testSet)

Runs evaluation procedure for given neural network and data set through all evaluatoors Evaluation results are stored in evaluators

Parameters:

neuralNetwork - trained neural network

testSet - test data set used for evaluation

Returns:

addEvaluator

public void addEvaluator(Evaluator evaluator)

Parameters:

evaluator -

getEvaluator

public <T extends Evaluator> T getEvaluator(Class<T> type)

Type Parameters:

T - Evaluator class

Parameters:

type - concrete evaluator class

Returns:

result of evaluation for given Evaluator type

getEvaluators

public Map<Class<?>,Evaluator> getEvaluators()

Return all evaluators used for evaluation

Returns:

getMeanSquareError

public double getMeanSquareError()

runFullEvaluation

public static void runFullEvaluation(NeuralNetwork<?> neuralNet,
                                     DataSet dataSet)

Out of the box method (util) which computes all metrics for given neural network and test data set

Parameters:

neuralNet - neural network to evaluate

dataSet - data set to evaluate