public void TanH2()
{
// ARRANGE
BackPropNetwork backprop = new BackPropNetwork();
List <double> inputs = new List <double> {
0.0,
0.7,
0.4,
0.1,
0.2222
};
// ACT
List <double> outputs = new List <double>();
foreach (var input in inputs)
{
outputs.Add(backprop.TanH(input));
}
// ASSERT
foreach (var output in outputs)
{
Assert.That(output, Is.InRange(0f, 1f));
}
}
public ImageCompress()
{
int[] layersizes = new int[3] {
8, 18, 8
};
ActivationFunction[] activFunctions = new ActivationFunction[3] {
ActivationFunction.None, ActivationFunction.Sigmoid, ActivationFunction.Linear
};
XmlDocument xdoc = new XmlDocument();
xdoc.Load(Server.MapPath("resources/ann.xml"));
ds = new DataSet();
ds.Load((XmlElement)xdoc.DocumentElement.ChildNodes[0]);
bpnetwork = new BackPropNetwork(layersizes, activFunctions);
nt = new NetworkTrainer(bpnetwork, ds);
nt.maxError = 0.00001;
nt.maxiterations = 10000;
nt.nudgewindow = 500;
nt.traininrate = 0.1;
nt.TrainDataset();
// save error
double[] err = nt.geteHistory();
string[] filedata = new string[err.Length];
for (int i = 0; i < err.Length; i++)
{
filedata[i] = i.ToString() + " " + err[i].ToString();
}
}
public void TanH1()
{
// ARRANGE
BackPropNetwork backprop = new BackPropNetwork();
List <double> inputs = new List <double> {
0.056566161,
0.3651,
0.268461,
0.00005,
0.00123005,
0.0,
0.0000125,
0.0440005,
1
};
// ACT
List <double> outputs = new List <double>();
foreach (var input in inputs)
{
outputs.Add(backprop.TanH(input));
}
// ASSERT
foreach (var output in outputs)
{
Assert.That(output, Is.InRange(0f, 1f));
}
}
//Restarts training process
public void RestartCurrent()
{
ResetCarPosition();
backProp = new BackPropNetwork();
Lap.lapCount = 0;
Rigidbody.GetComponent <CarPhysics>().driver = Driver.USER;
runOnce = true;
}
//Initialise objects
void Start()
{
Rigidbody = GetComponent <Rigidbody>();
backProp = new BackPropNetwork();
Rigidbody.GetComponent <CarPhysics>().driver = Driver.BackProp;
controller = new CarController();
saveLoad = new BackPropWeights();
Rigidbody.GetComponent <Sensor>().sensorLenght += 5;
position = new ResetPosition();
}
public void Random2()
{
// ARRANGE
BackPropNetwork backprop = new BackPropNetwork();
// ACT
double random = backprop.GetRandomWeight();
// ASSERT
Assert.That(random, Is.InRange(-1f, 1f));
}
//Initialise objects
void Start()
{
Rigidbody = GetComponent <Rigidbody>();
Reset.onClick.AddListener(ResetCurrent);
Restart.onClick.AddListener(RestartCurrent);
LearningRate = LearningModeScript.LearningRate;
NoLaps = LearningModeScript.NoLaps;
backProp = new BackPropNetwork();
Rigidbody.GetComponent <CarPhysics>().driver = Driver.USER;
controller = new CarController();
saveLoad = new BackPropWeights();
Rigidbody.GetComponent <Sensor>().sensorLenght += 5;
position = new ResetPosition();
}
public void FeedForward2()
{
// ARRANGE
BackPropNetwork backprop = new BackPropNetwork();
double[] inputs = { 0.6, 0.4, 0.2, 0, 0 };
int expectedlength = 2;
// ACT
double[] outputs = backprop.FeedForward(inputs);
// ASSERT
Assert.That(outputs.Length, Is.EqualTo(expectedlength));
foreach (var output in outputs)
{
Assert.That(output, Is.InRange(-1f, 1f));
}
}
public void Init1()
{
// ARRANGE
BackPropNetwork backprop = new BackPropNetwork();
int expectedlength = 3;
// ACT
// ASSERT
Assert.That(backprop.GetWeights().Count, Is.EqualTo(expectedlength));
Assert.That(backprop.weightsDer.Count, Is.EqualTo(expectedlength));
Assert.That(backprop.layerInputs.Count, Is.EqualTo(expectedlength));
Assert.That(backprop.layerOutputs.Count, Is.EqualTo(expectedlength));
Assert.That(backprop.layerLoss.Count, Is.EqualTo(expectedlength));
Assert.That(backprop.GetWeights().GetType() == typeof(List <double[][]>));
Assert.That(backprop.weightsDer.GetType() == typeof(List <double[][]>));
Assert.That(backprop.layerInputs.GetType() == typeof(List <double[]>));
Assert.That(backprop.layerOutputs.GetType() == typeof(List <double[]>));
Assert.That(backprop.layerLoss.GetType() == typeof(List <double[]>));
}
public CompressText()
{
int[] layersizes = new int[10] {
1, 10, 9, 8, 7, 5, 4, 3, 2, 1
};
ActivationFunction[] activFunctions = new ActivationFunction[10] {
ActivationFunction.None, ActivationFunction.Gaussian, ActivationFunction.Sigmoid, ActivationFunction.Sigmoid, ActivationFunction.Sigmoid, ActivationFunction.Sigmoid, ActivationFunction.Sigmoid, ActivationFunction.Sigmoid, ActivationFunction.Sigmoid,
ActivationFunction.Linear
};
XmlDocument xdoc = new XmlDocument();
xdoc.Load(Path.Combine(HttpRuntime.AppDomainAppPath, "resources/ann.xml"));
ds = new DataSet();
ds.Load((XmlElement)xdoc.DocumentElement.ChildNodes[0]);
bpnetwork = new BackPropNetwork(layersizes, activFunctions);
nt = new NetworkTrainer(bpnetwork, ds);
nt.maxError = 0.1;
nt.maxiterations = 10000;
nt.traininrate = 0.1;
nt.TrainDataset();
// save error
double[] err = nt.geteHistory();
string[] filedata = new string[err.Length];
for (int i = 0; i < err.Length; i++)
{
filedata[i] = i.ToString() + " " + err[i].ToString();
}
}