static void Main(string[] args)
{
string outputFile = "..\\..\\IrisTrainTestStats.txt";
FileStream ofs = new FileStream(outputFile, FileMode.Append, FileAccess.Write);
StreamWriter sw = new StreamWriter(ofs);
sw.WriteLine("\nBegin neural network training demo");
sw.WriteLine("\nData is the famous Iris flower set.");
sw.WriteLine("Predict species from sepal length, width, petal length, width");
sw.WriteLine("Iris setosa = 0 0 1, versicolor = 0 1 0, virginica = 1 0 0 \n");
sw.WriteLine("Raw data resembles:");
sw.WriteLine(" 5.1, 3.5, 1.4, 0.2, Iris setosa");
sw.WriteLine(" 7.0, 3.2, 4.7, 1.4, Iris versicolor");
sw.WriteLine(" 6.3, 3.3, 6.0, 2.5, Iris virginica");
sw.WriteLine(" ......\n");
double[][] allData = new double[150][];
string dataFile = "..\\..\\IrisData.txt";
allData = LoadData(dataFile, 150, 7);
sw.WriteLine("\nFirst 6 rows of the 150-item data set:");
ShowMatrix(sw, allData, 6, 1, true);
sw.WriteLine("Creating 80% training and 20% test data matrices");
double[][] trainData = null;
double[][] testData = null;
MakeTrainTest(allData, 72, out trainData, out testData); // seed = 72 gives a pretty demo.
sw.WriteLine("\nFirst 3 rows of training data:");
ShowMatrix(sw, trainData, 3, 1, true);
sw.WriteLine("First 3 rows of test data:");
ShowMatrix(sw, testData, 3, 1, true);
sw.WriteLine("\nCreating a 4-input, 7-hidden, 3-output neural network");
sw.Write("Hard-coded tanh function for input-to-hidden and softmax for ");
sw.WriteLine("hidden-to-output activations");
int numInput = 4;
int numHidden = 7;
int numOutput = 3;
NeuralNetwork nn = new NeuralNetwork(numInput, numHidden, numOutput);
int maxEpochs = 1000;
double learnRate = 0.05;
double momentum = 0.01;
sw.WriteLine("Setting maxEpochs = " + maxEpochs + ", learnRate = " + learnRate + ", momentum = " + momentum);
sw.WriteLine("Training has hard-coded mean squared " + "error < 0.040 stopping condition");
sw.WriteLine("\nBeginning training using incremental back-propagation\n");
nn.Train(trainData, maxEpochs, learnRate, momentum);
sw.WriteLine("Training complete");
double[] weights = nn.GetWeights();
sw.WriteLine("Final neural network weights and bias values:");
ShowVector(sw, weights, 10, 3, true);
double trainAcc = nn.Accuracy(trainData);
sw.WriteLine("\nAccuracy on training data = " + trainAcc.ToString("F4"));
double testAcc = nn.Accuracy(testData);
sw.WriteLine("\nAccuracy on test data = " + testAcc.ToString("F4"));
sw.WriteLine("\nEnd neural network training demo\n");
Console.ReadLine();
sw.Close();
ofs.Close();
} // Main