/// <summary>
/// Test cross validation
/// </summary>
public static void Test_CalcProbability_Xval(int data_set_index)
{
List <MLDataPoint> X = MLDataPointUtil.LoadDataSet(string.Format("X{0}.txt", data_set_index));
List <MLDataPoint> Xval = MLDataPointUtil.LoadDataSet(string.Format("Xval{0}.txt", data_set_index));
MultiVariateGaussianDistributionAD <MLDataPoint> algorithm = new MultiVariateGaussianDistributionAD <MLDataPoint>();
algorithm.ComputeGaussianDistribution(X);
List <MLDataPoint> correct_pval = MLDataPointUtil.LoadDataSet(string.Format("pval{0}.txt", data_set_index));
int row_count = Xval.Count;
double total_error = 0;
for (int i = 0; i < row_count; ++i)
{
double pval = algorithm.CalcProbability(Xval[i]);
double error = System.Math.Abs(correct_pval[i][0] - pval);
total_error += error;
Console.WriteLine("pval={0} correct_pval={1}", pval, correct_pval[i]);
}
Console.WriteLine("Total error: {0}", total_error);
}
public static void Test_FindOutliers(int data_set_index)
{
double threshold, F1Score;
SelectThreshold(data_set_index, out F1Score, out threshold);
List <MLDataPoint> X = MLDataPointUtil.LoadDataSet(string.Format("X{0}.txt", data_set_index));
List <MLDataPoint> correct_outliers = MLDataPointUtil.LoadDataSet(string.Format("outliers{0}.txt", data_set_index));
Console.WriteLine("Correct Outliers:");
for (int i = 0; i < correct_outliers.Count; ++i)
{
Console.WriteLine("{0}", correct_outliers[i][0]);
}
MultiVariateGaussianDistributionAD <MLDataPoint> algorithm = new MultiVariateGaussianDistributionAD <MLDataPoint>();
algorithm.ComputeGaussianDistribution(X);
List <int> outliers = algorithm.FindOutliers(X, threshold);
Console.WriteLine("Predict Outliers:");
for (int i = 0; i < outliers.Count; ++i)
{
Console.WriteLine("{0}", outliers[i] + 1);
}
Console.WriteLine("Point Count: {0}", X.Count);
Console.WriteLine("Threshold: {0}", threshold);
Console.WriteLine("Predict Outliers Count; {0} Correct Outliers Count: {1}", outliers.Count, correct_outliers.Count);
}