protected internal override void onStart()
{
base.onStart();
// set weights between input and rbf layer using kmeans
KMeansClustering kmeans = new KMeansClustering(TrainingSet);
kmeans.NumberOfClusters = neuralNetwork.getLayerAt(1).NeuronsCount; // set number of clusters as number of rbf neurons
kmeans.doClustering();
// get clusters (centroids)
Cluster[] clusters = kmeans.Clusters;
// assign each rbf neuron to one cluster
// and use centroid vectors to initialize neuron's input weights
Layer rbfLayer = neuralNetwork.getLayerAt(1);
int i = 0;
foreach (Neuron neuron in rbfLayer.Neurons)
{
KVector centroid = clusters[i].Centroid;
double[] weightValues = centroid.Values;
int c = 0;
foreach (Connection conn in neuron.InputConnections)
{
conn.Weight.Value = weightValues[c];
c++;
}
i++;
}
// get cluster centroids as list
List <KVector> centroids = new List <KVector>();
foreach (Cluster cluster in clusters)
{
centroids.Add(cluster.Centroid);
}
// use KNN to calculate sigma param - gausssian function width for each neuron
KNearestNeighbour knn = new KNearestNeighbour();
knn.DataSet = centroids;
int n = 0;
foreach (KVector centroid in centroids)
{
// calculate and set sigma for each neuron in rbf layer
KVector[] nearestNeighbours = knn.getKNearestNeighbours(centroid, k);
double sigma = calculateSigma(centroid, nearestNeighbours); // calculate in method
Neuron neuron = rbfLayer.getNeuronAt(n);
((Gaussian)neuron.TransferFunction).Sigma = sigma;
i++;
}
}
private void classifyBtn_Click(object sender, EventArgs e)
{
Bitmap currentImage = new Bitmap(pictureBox1.Image);
if (ClassifierType == "KNN")
{
KNearestNeighbour knnClassifier = new KNearestNeighbour(10, Main.trainingImagesFeatures, Main.trainingLabels);
pictureBox2.Image = pictureBox1.Image;
Bitmap drawnImage = new Bitmap(scaledImage((Bitmap)pictureBox1.Image));
pictureBox2.Image = drawnImage;
//byte[] imageFeatures = bitmapToBuffer(drawnImage);
// int classIndex = knnClassifier.classifySorting(K, imageFeatures);
// MessageBox.Show(classIndex.ToString());
}
}
private void button1_Click(object sender, EventArgs e)
{
dataGridView1.Rows.Clear();
dataGridView1.Refresh();
int index = 0;
string path = Path.Combine(Path.GetDirectoryName(Application.ExecutablePath), "Output.txt");
StreamWriter file = new StreamWriter(@path);
KNearestNeighbour knnClassifier = new KNearestNeighbour(10, Main.trainingImagesFeatures, Main.trainingLabels);
this.confusionMatrix = new int[10, 10];
var watch = Stopwatch.StartNew();
if (ClassifierType == "KNN")
{
while (index != NumberOfImages)
{
int classIndex = knnClassifier.classify(K, Mode.testImagesFeatures[index]);
this.confusionMatrix[Mode.testLabels[index], classIndex]++;
string[] output = { index.ToString(), Mode.testLabels[index].ToString(), classIndex.ToString() };
this.dataGridView1.Rows.Add(output);
string outputFile = "Image# " + index.ToString() + "\t Predection: "
+ classIndex.ToString() + "\t Label: " + Mode.testLabels[index] + "\n";
file.WriteLine(outputFile);
index++;
}
watch.Stop();
file.WriteLine(watch.ElapsedMilliseconds.ToString());
Process proc = Process.GetCurrentProcess();
file.WriteLine((proc.PrivateMemorySize64 / 1024).ToString());
}
else if (ClassifierType == "KNN Modified")
{
while (index != NumberOfImages)
{
int classIndex = knnClassifier.classifyModified(Mode.testImagesFeatures[index]);
this.confusionMatrix[Mode.testLabels[index], classIndex]++;
string[] output = { index.ToString(), Mode.testLabels[index].ToString(), classIndex.ToString() };
this.dataGridView1.Rows.Add(output);
string outputFile = "Image# " + index.ToString() + "\t Predection: "
+ classIndex.ToString() + "\t Label: " + Mode.testLabels[index] + "\n";
file.WriteLine(outputFile);
index++;
}
watch.Stop();
file.WriteLine(watch.ElapsedMilliseconds.ToString());
Process proc = Process.GetCurrentProcess();
file.WriteLine((proc.PrivateMemorySize64 / 1024).ToString());
}
file.Close();
ConfusionMatrixForm cmf = new ConfusionMatrixForm();
cmf.Show();
cmf.setConfusionMatrix(this.confusionMatrix, watch.ElapsedMilliseconds.ToString());
}
