public void LearningCurves_Calculate()
{
#region Read data
// Use StreamReader(filepath) when running from filesystem
var parser = new CsvParser(() => new StringReader(Resources.winequality_white));
var targetName = "quality";
// read feature matrix
var observations = parser.EnumerateRows(c => c != targetName)
.ToF64Matrix();
// read regression targets
var targets = parser.EnumerateRows(targetName)
.ToF64Vector();
#endregion
// metric for measuring model error
var metric = new MeanSquaredErrorRegressionMetric();
// creates cross validator, observations are shuffled randomly
var learningCurveCalculator = new RandomShuffleLearningCurvesCalculator <double>(metric,
samplePercentages: new double[] { 0.05, 0.1, 0.2, 0.4, 0.8, 1.0 },
trainingPercentage: 0.7, numberOfShufflesPrSample: 5);
// create learner
var learner = new RegressionDecisionTreeLearner(maximumTreeDepth: 5);
// calculate learning curve
var learningCurve = learningCurveCalculator.Calculate(learner, observations, targets);
// write to csv
var writer = new StringWriter();
learningCurve.Write(() => writer);
// trace result
// Plotting the learning curves will help determine if the model has high bias or high variance.
// This information can be used to determine what to try next in order to improve the model.
Trace.WriteLine(writer.ToString());
// alternatively, write to file
//learningCurve.Write(() => new StreamWriter(filePath));
}
public void RandomLearningCurvesCalculator_Calculate()
{
var sut = new RandomShuffleLearningCurvesCalculator <double>(
new MeanSquaredErrorRegressionMetric(),
new double[] { 0.2, 0.8 },
0.8, 42, 5);
var(observations, targets) = DataSetUtilities.LoadDecisionTreeDataSet();
var actual = sut.Calculate(new RegressionDecisionTreeLearner(),
observations, targets);
var expected = new List <LearningCurvePoint>()
{
new LearningCurvePoint(32, 0, 0.141565953928265),
new LearningCurvePoint(128, 0.0, 0.068970597423950036)
};
CollectionAssert.AreEqual(expected, actual);
}
public void RandomLearningCurvesCalculator_Calculate()
{
var sut = new RandomShuffleLearningCurvesCalculator <double>(new MeanSquaredErrorRegressionMetric(),
new double[] { 0.2, 0.8 }, 0.8, 42, 5);
var targetName = "T";
var parser = new CsvParser(() => new StringReader(Resources.DecisionTreeData));
var observations = parser.EnumerateRows(v => !v.Contains(targetName)).ToF64Matrix();
var targets = parser.EnumerateRows(targetName).ToF64Vector();
var actual = sut.Calculate(new RegressionDecisionTreeLearner(),
observations, targets);
var expected = new List <LearningCurvePoint>()
{
new LearningCurvePoint(32, 0, 0.141565953928265),
new LearningCurvePoint(128, 0.0, 0.068970597423950036)
};
CollectionAssert.AreEqual(expected, actual);
}