public ScoreModel(TrainModel trainedModel, int K = 3)
{
if (!trainedModel.Trained)
throw new Exception("TrainModel must be trained first");
this.trainedModel = trainedModel;
ConfussionMatrix = new int[trainedModel.NumberofClasses, trainedModel.NumberofClasses];
if (trainedModel.cType == ClassifierType.Bayesian)
classifier = new BayesianClassifier(trainedModel);
else if (trainedModel.cType == ClassifierType.KNearestNeighbour)
classifier = new KNearestNeighbour(trainedModel, K);
}
public ScoreModel(TrainModel trainedModel, int K = 3)
{
if (!trainedModel.Trained)
{
throw new Exception("TrainModel must be trained first");
}
this.trainedModel = trainedModel;
ConfussionMatrix = new int[trainedModel.NumberofClasses, trainedModel.NumberofClasses];
if (trainedModel.cType == ClassifierType.Bayesian)
{
classifier = new BayesianClassifier(trainedModel);
}
else if (trainedModel.cType == ClassifierType.KNearestNeighbour)
{
classifier = new KNearestNeighbour(trainedModel, K);
}
}
private void button2_Click(object sender, EventArgs e)
{
List<Data> Trainingdataset = Utility.LoadDataset(DatasetType.Training);
List<Data> Testingingdataset = Utility.LoadDataset(DatasetType.Testing);
TrainModel train = null;
if (comboBox1.SelectedIndex == 0)
train = new TrainModel(ViewModel.FeatureExtraction.FeatureExtractionType.EuclideanDistance, ClassifierType.Bayesian);
else if (comboBox1.SelectedIndex == 1)
train = new TrainModel(ViewModel.FeatureExtraction.FeatureExtractionType.EuclideanDistance, ClassifierType.KNearestNeighbour);
train.Train(Trainingdataset);
ScoreModel score = new ScoreModel(train, int.Parse(KNNTextBox.Text));
score.Score(Testingingdataset);
dataGridView1.Rows.Clear();
for (int i = 0; i < score.ConfussionMatrix.GetLength(0); ++i)
{
string[] arr = new string[score.ConfussionMatrix.GetLength(1)];
for (int j = 0; j < score.ConfussionMatrix.GetLength(1); ++j)
arr[j] = score.ConfussionMatrix[i, j].ToString();
dataGridView1.Rows.Add(arr);
}
OverallAccuracyLabel.Text = score.Accuracy.ToString();
}
private void ClassificationButton_Click(object sender, EventArgs e)
{
List<Data> Trainingdataset = Utility.LoadDataset(DatasetType.Training);
TrainModel train = null;
if (comboBox1.SelectedIndex == 0)
train = new TrainModel(ViewModel.FeatureExtraction.FeatureExtractionType.EuclideanDistance, ClassifierType.Bayesian);
else if (comboBox1.SelectedIndex == 1)
train = new TrainModel(ViewModel.FeatureExtraction.FeatureExtractionType.EuclideanDistance, ClassifierType.KNearestNeighbour);
train.Train(Trainingdataset);
OpenFileDialog f = new OpenFileDialog();
if (DialogResult.OK == f.ShowDialog())
{
ScoreModel score = new ScoreModel(train, int.Parse(KNNTextBox.Text));
List<Vector2> features = Utility.LoadPoints(f.FileName);
int estimatedClass = score.Classify(features);
ExpectedClassLabel.Text = estimatedClass.ToString();
AppModel app = new AppModel(@"C:\Program Files (x86)\Notepad++\notepad++.exe");
try
{
if (estimatedClass == 0)
app.Close();
else if (estimatedClass == 1)
app.Minimize();
else if (estimatedClass == 2)
app.Open();
else if (estimatedClass == 3)
app.Restore();
}
catch
{
MessageBox.Show("Couldn't take that action, please make sure that the specified program is already opened");
}
}
/*for(int i=0;i<40;)
{
Vector2 a = new Vector2();
a.x = double.Parse(inputarr[i++].Text);
a.y = double.Parse(inputarr[i++].Text);
features.Add(a);
}*/
}