public void TestExecute()
{
var mxKFoldRun = new MxKFoldRun(5, 2);
var experimentPerformance =
mxKFoldRun.Execute(new Classification.Experiment.Experiment(new C45(), new C45Parameter(1, true, 0.2), iris));
Assert.AreEqual(68.67, 100 * experimentPerformance.MeanPerformance().GetErrorRate(), 0.01);
experimentPerformance =
mxKFoldRun.Execute(new Classification.Experiment.Experiment(new C45(), new C45Parameter(1, true, 0.2), tictactoe));
Assert.AreEqual(84.66, 100 * experimentPerformance.MeanPerformance().GetErrorRate(), 0.01);
experimentPerformance =
mxKFoldRun.Execute(new Classification.Experiment.Experiment(new Knn(), new KnnParameter(1, 3, new EuclidianDistance()), bupa));
Assert.AreEqual(33.61, 100 * experimentPerformance.MeanPerformance().GetErrorRate(), 0.01);
experimentPerformance = mxKFoldRun.Execute(new Classification.Experiment.Experiment(new Knn(),
new KnnParameter(1, 3, new EuclidianDistance()), dermatology));
Assert.AreEqual(18.31, 100 * experimentPerformance.MeanPerformance().GetErrorRate(), 0.01);
experimentPerformance = mxKFoldRun.Execute(new Classification.Experiment.Experiment(new Lda(), new Parameter(1), bupa));
Assert.AreEqual(30.43, 100 * experimentPerformance.MeanPerformance().GetErrorRate(), 0.01);
experimentPerformance = mxKFoldRun.Execute(new Classification.Experiment.Experiment(new Lda(), new Parameter(1), dermatology));
Assert.AreEqual(5.46, 100 * experimentPerformance.MeanPerformance().GetErrorRate(), 0.01);
experimentPerformance = mxKFoldRun.Execute(new Classification.Experiment.Experiment(new LinearPerceptron(),
new LinearPerceptronParameter(1, 0.1, 0.99, 0.2, 100), iris));
Assert.AreEqual(82.67, 100 * experimentPerformance.MeanPerformance().GetErrorRate(), 0.01);
experimentPerformance = mxKFoldRun.Execute(new Classification.Experiment.Experiment(new LinearPerceptron(),
new LinearPerceptronParameter(1, 0.1, 0.99, 0.2, 100), dermatology));
Assert.AreEqual(7.92, 100 * experimentPerformance.MeanPerformance().GetErrorRate(), 0.01);
experimentPerformance = mxKFoldRun.Execute(new Classification.Experiment.Experiment(new NaiveBayes(), new Parameter(1), car));
Assert.AreEqual(34.09, 100 * experimentPerformance.MeanPerformance().GetErrorRate(), 0.01);
experimentPerformance = mxKFoldRun.Execute(new Classification.Experiment.Experiment(new NaiveBayes(), new Parameter(1), nursery));
Assert.AreEqual(38.97, 100 * experimentPerformance.MeanPerformance().GetErrorRate(), 0.01);
experimentPerformance =
mxKFoldRun.Execute(new Classification.Experiment.Experiment(new Bagging(), new BaggingParameter(1, 50), tictactoe));
Assert.AreEqual(84.55, 100 * experimentPerformance.MeanPerformance().GetErrorRate(), 0.01);
experimentPerformance = mxKFoldRun.Execute(new Classification.Experiment.Experiment(new Bagging(), new BaggingParameter(1, 50), car));
Assert.AreEqual(34.20, 100 * experimentPerformance.MeanPerformance().GetErrorRate(), 0.01);
experimentPerformance = mxKFoldRun.Execute(new Classification.Experiment.Experiment(new Dummy(), new Parameter(1), nursery));
Assert.AreEqual(84.57, 100 * experimentPerformance.MeanPerformance().GetErrorRate(), 0.01);
experimentPerformance = mxKFoldRun.Execute(new Classification.Experiment.Experiment(new Dummy(), new Parameter(1), iris));
Assert.AreEqual(100.0, 100 * experimentPerformance.MeanPerformance().GetErrorRate(), 0.01);
}
public void TestCompare()
{
var mxKFoldRun = new MxKFoldRun(5, 2);
var experimentPerformance1 =
mxKFoldRun.Execute(new Classification.Experiment.Experiment(new C45(), new C45Parameter(1, true, 0.2), iris));
var experimentPerformance2 = mxKFoldRun.Execute(new Classification.Experiment.Experiment(new LinearPerceptron(),
new LinearPerceptronParameter(1, 0.1, 0.99, 0.2, 100), iris));
var paired5x2t = new Paired5x2t();
Assert.AreEqual(0.329, paired5x2t.Compare(experimentPerformance1, experimentPerformance2).GetPValue(), 0.001);
experimentPerformance1 =
mxKFoldRun.Execute(new Classification.Experiment.Experiment(new C45(), new C45Parameter(1, true, 0.2), tictactoe));
experimentPerformance2 =
mxKFoldRun.Execute(new Classification.Experiment.Experiment(new Bagging(), new BaggingParameter(1, 50), tictactoe));
Assert.AreEqual(0.00159, paired5x2t.Compare(experimentPerformance1, experimentPerformance2).GetPValue(),
0.00001);
experimentPerformance1 = mxKFoldRun.Execute(new Classification.Experiment.Experiment(new Lda(), new Parameter(1), dermatology));
experimentPerformance2 = mxKFoldRun.Execute(new Classification.Experiment.Experiment(new LinearPerceptron(),
new LinearPerceptronParameter(1, 0.1, 0.99, 0.2, 100), dermatology));
Assert.AreEqual(0.769, paired5x2t.Compare(experimentPerformance1, experimentPerformance2).GetPValue(),
0.001);
experimentPerformance1 = mxKFoldRun.Execute(new Classification.Experiment.Experiment(new Dummy(), new Parameter(1), nursery));
experimentPerformance2 = mxKFoldRun.Execute(new Classification.Experiment.Experiment(new NaiveBayes(), new Parameter(1), nursery));
Assert.AreEqual(0.0, paired5x2t.Compare(experimentPerformance1, experimentPerformance2).GetPValue(), 0.0001);
experimentPerformance1 = mxKFoldRun.Execute(new Classification.Experiment.Experiment(new NaiveBayes(), new Parameter(1), car));
experimentPerformance2 =
mxKFoldRun.Execute(new Classification.Experiment.Experiment(new Bagging(), new BaggingParameter(1, 50), car));
Assert.AreEqual(0.0078, paired5x2t.Compare(experimentPerformance1, experimentPerformance2).GetPValue(),
0.0001);
experimentPerformance1 =
mxKFoldRun.Execute(new Classification.Experiment.Experiment(new Knn(), new KnnParameter(1, 3, new EuclidianDistance()), bupa));
experimentPerformance2 = mxKFoldRun.Execute(new Classification.Experiment.Experiment(new Lda(), new Parameter(1), bupa));
Assert.AreEqual(0.035, paired5x2t.Compare(experimentPerformance1, experimentPerformance2).GetPValue(), 0.001);
}