public void ApplyCrossValidation(int folds, double[][] inputs, int[] outputs)
{
crossValidation = CrossValidation.Create(
k: 10, // We will be using 10-fold cross validation
learner: (p) => new B.NaiveBayesLearning() // here we create the learning algorithm
{
},
// Now we have to specify how the tree performance should be measured:
loss: (actual, expected, p) => new ZeroOneLoss(expected).Loss(actual),
// This function can be used to perform any special
// operations before the actual learning is done, but
// here we will just leave it as simple as it can be:
fit: (teacher, x, y, w) => teacher.Learn(x, y, w),
// Finally, we have to pass the input and output data
// that will be used in cross-validation.
x: inputs.Select(v => v.Select(u => Convert.ToInt32(u)).ToArray()).ToArray(), y: outputs
);
var result = crossValidation.Learn(inputs.Select(u => u.Select(v => Convert.ToInt32(v)).ToArray()).ToArray(), outputs);
ConfusionMatrix = result.ToConfusionMatrix(inputs.Select(v => v.Select(u => Convert.ToInt32(u)).ToArray()).ToArray(), outputs).Matrix;
}
public static void SVM_crossValidation_AnomalyClassifier(double[][] inputs, int[] outputs)
{
Accord.Math.Random.Generator.Seed = 0;
var crossvalidation = new CrossValidation <SupportVectorMachine <Gaussian, double[]>, double[]>()
{
K = 10, // Use 3 folds in cross-validation
//Learner = (s) => new SequentialMinimalOptimization<Gaussian, double[]>()
//{
// Complexity = 10,
// Kernel = new Gaussian(3)
//},
Learner = (s) => new SequentialMinimalOptimization <Gaussian, double[]>()
{
Complexity = 10,
Kernel = new Gaussian(2.6)
},
// Indicate how the performance of those models will be measured
Loss = (expected, actual, p) => new ZeroOneLoss(expected).Loss(actual),
Stratify = true, // force balancing of classes
};
// If needed, control the parallelization degree
crossvalidation.ParallelOptions.MaxDegreeOfParallelism = 1;
// Compute the cross-validation
var result = crossvalidation.Learn(inputs, outputs);
// Finally, access the measured performance.
double trainingErrors = result.Training.Mean;
double validationErrors = result.Validation.Mean;
// If desired, compute an aggregate confusion matrix for the validation sets:
GeneralConfusionMatrix gcm = result.ToConfusionMatrix(inputs, outputs);
double accuracy = gcm.Accuracy;
double error = gcm.Error;
Console.WriteLine("Accuracy:" + gcm.Accuracy);
Console.WriteLine("Error:" + gcm.Error);
Console.WriteLine("First Anomaly type Precision:" + gcm.Precision[0]);
Console.WriteLine("First Anomaly type Recall:" + gcm.Recall[0]);
Console.WriteLine("Second Anomaly type Precision:" + gcm.Precision[1]);
Console.WriteLine("Second Anomaly type Recall:" + gcm.Recall[1]);
//double anomalyFScore = 2 * (gcm.Precision[1] * gcm.Recall[1]) / (gcm.Precision[1] + gcm.Recall[1]);
//double NotAnomalyFScore = 2 * (gcm.Precision[0] * gcm.Recall[0]) / (gcm.Precision[0] + gcm.Recall[0]);
//Console.WriteLine("Not ANomaly F-score:" + NotAnomalyFScore);
//Console.WriteLine("Anomaly F-score:" + anomalyFScore);
Console.WriteLine("First Anomaly type F-score:" + gcm.PerClassMatrices[0].FScore);
Console.WriteLine("Second Anomaly type F-score:" + gcm.PerClassMatrices[1].FScore);
//var splitset = new SplitSetValidation<MulticlassSupportVectorMachine<Gaussian, double[]>, double[]>()
//{
// Learner = (s) => new MulticlassSupportVectorLearning<Gaussian, double[]>()
// {
// Learner = (m) => new SequentialMinimalOptimization<Gaussian, double[]>()
// {
// Complexity = 10,
// Kernel = new Gaussian(2.92)
// }
// }
// ,Stratify =true
//};
//var result = splitset.Learn(inputs, outputs);
//double trainingErrors = result.Training.Value;
//double validationErrors = result.Validation.Value;
}
// public static void NaiveBaise(double[][] chain)
// {
// DataTable data = new DataTable("dataa");
// data.Rows.Add()
// data.Columns.AddRange( "speedAvg1", "speedAvg2", "speedAvg3", "speedAvg4", "speedAvg5", "speedAvg6", "ContainAnomaly");
// Codification codebook = new Codification(data,
//"speedAvg1", "speedAvg2", "speedAvg3", "speedAvg4", "speedAvg5", "speedAvg6", "ContainAnomaly");
// // Extract input and output pairs to train
// DataTable symbols = codebook.Apply(data);
// int[][] inputs = symbols.ToArray<int>("Outlook", "Temperature", "Humidity", "Wind");
// int[] outputs = symbols.ToArray<int>("PlayTennis");
// }
public static void SVM_crossValidation_AnomalyDetection(double[][] inputs, int[] outputs)
{
Accord.Math.Random.Generator.Seed = 0;
var crossvalidation = new CrossValidation <SupportVectorMachine <Gaussian, double[]>, double[]>()
{
K = 100, // Use 3 folds in cross-validation
// Indicate how learning algorithms for the models should be created
//Learner = (s) => new SequentialMinimalOptimization<Gaussian, double[]>()
//{
// Complexity = 10,
// Kernel = new Gaussian(0.5)
//},
//if ssegments beside hyper removed
//Learner = (s) => new SequentialMinimalOptimization<Gaussian, double[]>()
//{
// Complexity = 100,
// Kernel = new Gaussian(1)
//},
//with max and min speed variance
//Learner = (s) => new SequentialMinimalOptimization<Gaussian, double[]>()
//{
// Complexity = 100,
// Kernel = new Gaussian(1.55)
//},
//Learner = (s) => new SequentialMinimalOptimization<Gaussian, double[]>()
//{
// Complexity = 15,
// Kernel = new Gaussian(1.55)
//},
//Learner = (s) => new SequentialMinimalOptimization<Gaussian, double[]>()
//{
// Complexity = 7,
// Kernel = new Gaussian(1.9)
//},
//best results for 7 segments witrh fields :sc.MaxSpeed.Value, sc.MinSpeed.Value, sc.SpeedRange.Value, sc.TotalTime.Value,sc.MinTime.Value,sc.MaxTime.Value, sc.TimeRange.Value,sc.MaxSpeedVar.Value,sc.MinSpeedVar.Value, sc.SpeedVarRange.Value
//Learner = (s) => new SequentialMinimalOptimization<Gaussian, double[]>()
//{
// Complexity = 7,
// Kernel = new Gaussian(2.4)
//},
//Best of the best
//Learner = (s) => new SequentialMinimalOptimization<Gaussian, double[]>()
//{
// Complexity = 7,
// Kernel = new Gaussian(2.4)
//},
//for Roadscanner DB
//Learner = (s) => new SequentialMinimalOptimization<Gaussian, double[]>()
//{
// Complexity = 10,
// Kernel = new Gaussian(2.92)
//},
//for Roadscanner3 DB
//without pothole
//Learner = (s) => new SequentialMinimalOptimization<Gaussian, double[]>()
//{
// Complexity = 10,
// Kernel = new Gaussian(2.1)
//},
//for Roadscanner3 DB
//with pothole
//Learner = (s) => new SequentialMinimalOptimization<Gaussian, double[]>()
//{
// Complexity = 15,
// Kernel = new Gaussian(2.5)
//},
//RoadScanner3 Db
//with pothole
//
Learner = (s) => new SequentialMinimalOptimization <Gaussian, double[]>()
{
Complexity = 15,
Kernel = new Gaussian(3)
},
// Indicate how the performance of those models will be measured
Loss = (expected, actual, p) => new ZeroOneLoss(expected).Loss(actual),
Stratify = true, // force balancing of classes
};
// If needed, control the parallelization degree
crossvalidation.ParallelOptions.MaxDegreeOfParallelism = 1;
// Compute the cross-validation
var result = crossvalidation.Learn(inputs, outputs);
// Finally, access the measured performance.
double trainingErrors = result.Training.Mean;
double validationErrors = result.Validation.Mean;
// If desired, compute an aggregate confusion matrix for the validation sets:
GeneralConfusionMatrix gcm = result.ToConfusionMatrix(inputs, outputs);
double accuracy = gcm.Accuracy;
double error = gcm.Error;
Console.WriteLine("Accuracy:" + gcm.Accuracy);
Console.WriteLine("Error:" + gcm.Error);
Console.WriteLine("Not Anomaly Precision:" + gcm.Precision[0]);
Console.WriteLine("Not Anomaly Recall:" + gcm.Recall[0]);
Console.WriteLine("Anomaly Precision:" + gcm.Precision[1]);
Console.WriteLine("Anomaly Recall:" + gcm.Recall[1]);
//double anomalyFScore = 2 * (gcm.Precision[1] * gcm.Recall[1]) / (gcm.Precision[1] + gcm.Recall[1]);
//double NotAnomalyFScore = 2 * (gcm.Precision[0] * gcm.Recall[0]) / (gcm.Precision[0] + gcm.Recall[0]);
//Console.WriteLine("Not ANomaly F-score:" + NotAnomalyFScore);
//Console.WriteLine("Anomaly F-score:" + anomalyFScore);
Console.WriteLine("Not ANomaly F-score:" + gcm.PerClassMatrices[0].FScore);
Console.WriteLine("Anomaly F-score:" + gcm.PerClassMatrices[1].FScore);
}
public void learn_hmm()
{
#region doc_learn_hmm
// Ensure results are reproducible
Accord.Math.Random.Generator.Seed = 0;
// This is a sample code on how to use Cross-Validation
// to assess the performance of Hidden Markov Models.
// Declare some testing data
int[][] inputs = new int[][]
{
new int[] { 0, 1, 1, 0 }, // Class 0
new int[] { 0, 0, 1, 0 }, // Class 0
new int[] { 0, 1, 1, 1, 0 }, // Class 0
new int[] { 0, 1, 1, 1, 0 }, // Class 0
new int[] { 0, 1, 1, 0 }, // Class 0
new int[] { 0, 0, 0, 0, 0 }, // Class 1
new int[] { 0, 0, 0, 1, 0 }, // Class 1
new int[] { 0, 0, 0, 0, 0 }, // Class 1
new int[] { 0, 0, 0 }, // Class 1
new int[] { 0, 0, 0, 0 }, // Class 1
new int[] { 1, 0, 0, 1 }, // Class 2
new int[] { 1, 1, 0, 1 }, // Class 2
new int[] { 1, 0, 0, 0, 1 }, // Class 2
new int[] { 1, 0, 1 }, // Class 2
new int[] { 1, 1, 0, 1 }, // Class 2
};
int[] outputs = new int[]
{
0, 0, 0, 0, 0, // First 5 sequences are of class 0
1, 1, 1, 1, 1, // Middle 5 sequences are of class 1
2, 2, 2, 2, 2, // Last 5 sequences are of class 2
};
// Create a new Cross-validation algorithm passing the data set size and the number of folds
var crossvalidation = new CrossValidation <HiddenMarkovClassifier, int[]>()
{
K = 3, // Use 3 folds in cross-validation
Learner = (s) => new HiddenMarkovClassifierLearning()
{
Learner = (p) => new BaumWelchLearning()
{
NumberOfStates = 3
}
},
Loss = (expected, actual, p) =>
{
var cm = new GeneralConfusionMatrix(classes: p.Model.NumberOfClasses, expected: expected, predicted: actual);
p.Variance = cm.Variance;
return(p.Value = cm.Kappa);
},
Stratify = false,
};
// If needed, control the parallelization degree
crossvalidation.ParallelOptions.MaxDegreeOfParallelism = 1;
// Compute the cross-validation
var result = crossvalidation.Learn(inputs, outputs);
// Finally, access the measured performance.
double trainingErrors = result.Training.Mean;
double validationErrors = result.Validation.Mean;
double trainingErrorVar = result.Training.Variance;
double validationErrorVar = result.Validation.Variance;
double trainingErrorPooledVar = result.Training.PooledVariance;
double validationErrorPooledVar = result.Validation.PooledVariance;
#endregion
Assert.AreEqual(5, result.AverageNumberOfSamples);
Assert.AreEqual(15, result.NumberOfSamples);
Assert.AreEqual(0, result.NumberOfInputs);
Assert.AreEqual(3, result.NumberOfOutputs);
Assert.AreEqual(3, result.Models.Length);
for (int i = 0; i < 3; i++)
{
Assert.AreEqual(0, result.Models[i].NumberOfInputs);
Assert.AreEqual(3, result.Models[i].NumberOfOutputs);
Assert.AreEqual(3, result.Models[i].Model.NumberOfClasses);
Assert.AreEqual(0, result.Models[i].Model.NumberOfInputs);
Assert.AreEqual(3, result.Models[i].Model.NumberOfOutputs);
Assert.AreEqual(10, result.Models[i].Training.NumberOfSamples);
Assert.AreEqual(5, result.Models[i].Validation.NumberOfSamples);
}
Assert.AreEqual(3, crossvalidation.K);
Assert.AreEqual(1.0, result.Training.Mean, 1e-10);
Assert.AreEqual(0.78472222222222232, result.Validation.Mean, 1e-10);
Assert.AreEqual(0, result.Training.Variance, 1e-10);
Assert.AreEqual(0.034866898148148126, result.Validation.Variance, 1e-10);
Assert.AreEqual(0, result.Training.StandardDeviation, 1e-10);
Assert.AreEqual(0.18672680082984372, result.Validation.StandardDeviation, 1e-10);
Assert.AreEqual(0, result.Training.PooledVariance);
Assert.AreEqual(0.045256528501157391, result.Validation.PooledVariance);
Assert.AreEqual(0, result.Training.PooledStandardDeviation);
Assert.AreEqual(0.21273581856649668, result.Validation.PooledStandardDeviation);
Assert.AreEqual(3, crossvalidation.Folds.Length);
Assert.AreEqual(3, result.Models.Length);
}
public void learn_test()
{
#region doc_learn
// Ensure results are reproducible
Accord.Math.Random.Generator.Seed = 0;
// This is a sample code on how to use Cross-Validation
// to assess the performance of Support Vector Machines.
// Consider the example binary data. We will be trying
// to learn a XOR problem and see how well does SVMs
// perform on this data.
double[][] data =
{
new double[] { -1, -1 }, new double[] { 1, -1 },
new double[] { -1, 1 }, new double[] { 1, 1 },
new double[] { -1, -1 }, new double[] { 1, -1 },
new double[] { -1, 1 }, new double[] { 1, 1 },
new double[] { -1, -1 }, new double[] { 1, -1 },
new double[] { -1, 1 }, new double[] { 1, 1 },
new double[] { -1, -1 }, new double[] { 1, -1 },
new double[] { -1, 1 }, new double[] { 1, 1 },
};
int[] xor = // result of xor for the sample input data
{
-1, 1,
1, -1,
-1, 1,
1, -1,
-1, 1,
1, -1,
-1, 1,
1, -1,
};
// Create a new Cross-validation algorithm passing the data set size and the number of folds
var crossvalidation = new CrossValidation <SupportVectorMachine <Linear, double[]>, double[]>()
{
K = 3, // Use 3 folds in cross-validation
// Indicate how learning algorithms for the models should be created
Learner = (s) => new SequentialMinimalOptimization <Linear, double[]>()
{
Complexity = 100
},
// Indicate how the performance of those models will be measured
Loss = (expected, actual, p) => new ZeroOneLoss(expected).Loss(actual),
Stratify = false, // do not force balancing of classes
};
// If needed, control the parallelization degree
crossvalidation.ParallelOptions.MaxDegreeOfParallelism = 1;
// Compute the cross-validation
var result = crossvalidation.Learn(data, xor);
// Finally, access the measured performance.
double trainingErrors = result.Training.Mean;
double validationErrors = result.Validation.Mean;
#endregion
Assert.AreEqual(3, crossvalidation.K);
Assert.AreEqual(0.37575757575757579, result.Training.Mean, 1e-10);
Assert.AreEqual(0.75555555555555554, result.Validation.Mean, 1e-10);
Assert.AreEqual(0.00044077134986225924, result.Training.Variance, 1e-10);
Assert.AreEqual(0.0059259259259259334, result.Validation.Variance, 1e-10);
Assert.AreEqual(0.020994555243259126, result.Training.StandardDeviation, 1e-10);
Assert.AreEqual(0.076980035891950155, result.Validation.StandardDeviation, 1e-10);
Assert.AreEqual(0, result.Training.PooledStandardDeviation);
Assert.AreEqual(0, result.Validation.PooledStandardDeviation);
Assert.AreEqual(3, crossvalidation.Folds.Length);
Assert.AreEqual(3, result.Models.Length);
}
public void learn_test()
{
#region doc_learn
// Ensure results are reproducible
Accord.Math.Random.Generator.Seed = 0;
// This is a sample code on how to use Cross-Validation
// to assess the performance of Support Vector Machines.
// Consider the example binary data. We will be trying
// to learn a XOR problem and see how well does SVMs
// perform on this data.
double[][] data =
{
new double[] { -1, -1 }, new double[] { 1, -1 },
new double[] { -1, 1 }, new double[] { 1, 1 },
new double[] { -1, -1 }, new double[] { 1, -1 },
new double[] { -1, 1 }, new double[] { 1, 1 },
new double[] { -1, -1 }, new double[] { 1, -1 },
new double[] { -1, 1 }, new double[] { 1, 1 },
new double[] { -1, -1 }, new double[] { 1, -1 },
new double[] { -1, 1 }, new double[] { 1, 1 },
};
int[] xor = // result of xor for the sample input data
{
-1, 1,
1, -1,
-1, 1,
1, -1,
-1, 1,
1, -1,
-1, 1,
1, -1,
};
// Create a new Cross-validation algorithm passing the data set size and the number of folds
var crossvalidation = new CrossValidation <SupportVectorMachine <Linear, double[]>, double[]>()
{
K = 3, // Use 3 folds in cross-validation
// Indicate how learning algorithms for the models should be created
Learner = (s) => new SequentialMinimalOptimization <Linear, double[]>()
{
Complexity = 100
},
// Indicate how the performance of those models will be measured
Loss = (expected, actual, p) => new ZeroOneLoss(expected).Loss(actual),
Stratify = false, // do not force balancing of classes
};
// If needed, control the parallelization degree
crossvalidation.ParallelOptions.MaxDegreeOfParallelism = 1;
// Compute the cross-validation
var result = crossvalidation.Learn(data, xor);
// Finally, access the measured performance.
double trainingErrors = result.Training.Mean; // should be 0.30606060606060609 (+/- var. 0.083498622589531682)
double validationErrors = result.Validation.Mean; // should be 0.3666666666666667 (+/- var. 0.023333333333333334)
// If desired, compute an aggregate confusion matrix for the validation sets:
GeneralConfusionMatrix gcm = result.ToConfusionMatrix(data, xor);
double accuracy = gcm.Accuracy; // should be 0.625
double error = gcm.Error; // should be 0.375
#endregion
Assert.AreEqual(16, gcm.Samples);
Assert.AreEqual(0.625, accuracy);
Assert.AreEqual(0.375, error);
Assert.AreEqual(2, gcm.Classes);
Assert.AreEqual(3, crossvalidation.K);
Assert.AreEqual(0.30606060606060609, result.Training.Mean, 1e-10);
Assert.AreEqual(0.3666666666666667, result.Validation.Mean, 1e-10);
Assert.AreEqual(0.083498622589531682, result.Training.Variance, 1e-10);
Assert.AreEqual(0.023333333333333334, result.Validation.Variance, 1e-10);
Assert.AreEqual(0.28896128216342704, result.Training.StandardDeviation, 1e-10);
Assert.AreEqual(0.15275252316519467, result.Validation.StandardDeviation, 1e-10);
Assert.AreEqual(0, result.Training.PooledStandardDeviation);
Assert.AreEqual(0, result.Validation.PooledStandardDeviation);
Assert.AreEqual(3, crossvalidation.Folds.Length);
Assert.AreEqual(3, result.Models.Length);
}