static void Main(string[] args)
{
var mnistDataParser = new MnistDataParser();
var trainingRecords = mnistDataParser.Parse("d:\\train-images", "d:\\train-labels");
var testRecords = mnistDataParser.Parse("d:\\test-images", "d:\\test-labels");
var neuralNetwork = new Network(new[] {784, 30, 10});
var benchmark = new Benchmark(neuralNetwork);
for (double accuracy = 0; accuracy < .9;)
{
accuracy = benchmark.AccuracyFor(testRecords);
Console.WriteLine("Success rate: {0} %", accuracy*100);
neuralNetwork.Train(trainingRecords, .01, 10);
}
}
static void Main(string[] args)
{
var mnistDataParser = new MnistDataParser();
var trainingRecords = mnistDataParser.Parse("d:\\train-images", "d:\\train-labels");
var testRecords = mnistDataParser.Parse("d:\\test-images", "d:\\test-labels");
var neuralNetwork = new Network(new[] { 784, 30, 10 });
var benchmark = new Benchmark(neuralNetwork);
for (double accuracy = 0; accuracy < .9;)
{
accuracy = benchmark.AccuracyFor(testRecords);
Console.WriteLine("Success rate: {0} %", accuracy * 100);
neuralNetwork.Train(trainingRecords, .01, 10);
}
}