org.neuroph.eval.classification

Class McNemarTest

java.lang.Object

org.neuroph.eval.classification.McNemarTest

public class McNemarTest
extends java.lang.Object

Class used to test if there is statistical significant difference between two classifiers

Constructor Summary

Constructors

Constructor and Description
McNemarTest()

Method Summary

Modifier and Type	Method and Description
boolean	evaluateNetworks(NeuralNetwork network1, NeuralNetwork network2, DataSet dataSet)

Methods inherited from class java.lang.Object

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

Constructor Detail

McNemarTest

public McNemarTest()

Method Detail

evaluateNetworks

public boolean evaluateNetworks(NeuralNetwork network1,
                                NeuralNetwork network2,
                                DataSet dataSet)

Parameters:

network1 - first trained neurl netowrk

network2 - second trained neural network

dataSet - data set used for performance evaluation

Returns:

if there exists significant difference between two classification models