// Start is called before the first frame update
void Start()
{
// Init neural network which contains 2 inputs, 2 hidden layers
// (2 neurons each) and 1 output. It also uses Unity's random
// function and step function (every neuron's output value is
// 1 or 0)
net = new NeuralNet.NeuralNetwork(2, new int[2] {
2, 2
}, 1, Random.Range, ActivationFunctions.Step);
}
public Genome(int NumInputs, int NumOutputs, int NeuronsPerHiddenLayer, int NumHiddenLayers)
{
Net = new NeuralNetwork();
Net.NumInputs = NumInputs;
Net.NumOutputs = NumOutputs;
Net.NeuronsPerHiddenLayer = NeuronsPerHiddenLayer;
Net.NumHiddenLayers = NumHiddenLayers;
Fitness = 0;
}
/// <summary>
/// Machine learning attempt 3
/// </summary>
public static void Main(String[] args)
{
NeuralNetwork nn = new NeuralNetwork(new UInt16[] { 1089, 16, 24, 4 }) /*{descent=Program.learnRate}*/;
DateTime start;
TimeSpan ts;
List <Byte> answers;
String[] dirs = Directory.GetDirectories(@".\Dataset\"), files;
Byte[] desiredAnswer;
Random r = new Random();
String dir, file;
UInt32 correct = 0, incorrect = 0;
Boolean step = false, vis = false;
/*
* foreach (Byte @byte in
* nn.makePrediction(Util.imageToNeuralData(@"C:\Users\Elite\Documents\SharpDevelop Projects\NeuralNet\NeuralNet\bin\Debug\Dataset\0\number-1.png"),
* new Byte[]{255,0,0,0,0,0,0,0,0,0},
* false)
* )
* Console.WriteLine(@byte.ToString()+',');
*
* goto halt;
*/
trainSubRt:
while (!(Console.KeyAvailable && Console.ReadKey(true).Key == ConsoleKey.Escape))
{
start = DateTime.UtcNow;
desiredAnswer = new Byte[] { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 };
dir = (dirs[r.Next(0, 4)]);
files = Directory.GetFiles(dir);
file = files[r.Next(0, files.Length)];
desiredAnswer[Byte.Parse(dir.Last().ToString())] = 255;
// desiredAnswer=new Byte[]{255,0,0,0,0,0,0,0,0,0};
// file=@"C:\Users\Elite\Documents\SharpDevelop Projects\NeuralNet\NeuralNet\bin\Debug\Dataset\0\0.png";
answers = nn.makePrediction(Util.imageToNeuralData(file), desiredAnswer, true, vis, delegate(){ Program.Main(args); });
// foreach (Byte @byte in answers)
// Console.WriteLine(@byte.ToString()+',');
ts = DateTime.UtcNow - start;
Byte prediction = (Byte)(answers.IndexOf(answers.Max())), desiredPrediction = (Byte)(desiredAnswer.ToList().IndexOf(desiredAnswer.Max()));
if (prediction == desiredPrediction)
{
++correct;
}
else
{
++incorrect;
}
Console.WriteLine("Answer: " + prediction.ToString() + ", desired: " + desiredPrediction.ToString() + ", R/W: " + correct.ToString() + '/' + incorrect.ToString() + "(" + ts.TotalMilliseconds.ToString() + "ms, " + file + ')');
/*UInt32 total=correct+incorrect;
*
* if ((correct>incorrect&&((total)>8))||((total)>13000)) {
*
* Console.WriteLine("HL size:"+Program.hlSize.ToString());
*
* if (Program.hlSize==1) {
*
* List<String> lines=new List<String>();
* foreach (KeyValuePair<UInt16,List<UInt32>> kvp in Program.times) {
* String str="HL size: "+kvp.Key.ToString()+",Iterations (avg:"+kvp.Value.Select(x=>Convert.ToInt32(x)).Average().ToString()+"): ";
* foreach (UInt32 num in kvp.Value)
* str+=num.ToString()+',';
* lines.Add(str);
* }
*
* File.WriteAllLines("./results",lines);
* return;
*
* }
*
* if (Program.loopCtr==Program.loopCtrMax)
* Program.times.Add(Program.hlSize,new List<UInt32>(Program.loopCtrMax));
*
* times[Program.hlSize].Add(total);
*
* if (Program.loopCtr==0) {
*
* --Program.hlSize;
* Program.loopCtr=Program.loopCtrMax;
*
* }
* else--Program.loopCtr;
*
* Program.Main(args);
* return;
*
* }*/
if (step && Console.ReadKey().Key == ConsoleKey.Escape)
{
break;
}
}
while (true)
{
String str = Console.ReadLine();
if (str == "continue" || str == "train")
{
goto trainSubRt;
}
else if (str == "break" || str == "stop")
{
goto halt;
}
else if (str == "visual" || str == "visualize")
{
vis = !vis;
goto trainSubRt;
}
else if (str == "step")
{
step = !step;
goto trainSubRt;
}
else
{
List <Byte> results
= nn.makePrediction(Util.imageToNeuralData(str),
null,
false,
vis).ToList();
foreach (Byte @byte in results)
{
Console.Write(@byte.ToString() + ',');
}
Console.WriteLine("Your number was a: " + results.IndexOf(results.Max()).ToString());
}
}
halt:
goto halt;
}
public static NeuralNetwork LoadNetworkFromFile(String filename)
{
NeuralNetwork neuralNetwork = null;
try
{
StreamReader sr = new StreamReader(filename);
String str = sr.ReadLine(); // headers
String[] strTab;
str = sr.ReadLine();
strTab = str.Split(';');
bool isNetworkUsingBias = bool.Parse(strTab[1]);
str = sr.ReadLine();
strTab = str.Split(';');
double minWeight = double.Parse(strTab[1]);
str = sr.ReadLine();
strTab = str.Split(';');
double maxWeight = double.Parse(strTab[1]);
str = sr.ReadLine();
strTab = str.Split(';');
LearningMethod methodOfLearning = strTab[1] == "0" ? LearningMethod.LINEAR : LearningMethod.NOT_LINEAR;
str = sr.ReadLine();
strTab = str.Split(';');
String name = strTab[1];
String[] layersNeuronsStr = sr.ReadLine().Split(';'); // neurons in layers
String[] layersActivationFunctionStr = sr.ReadLine().Split(';'); // neurons in layers
List <LayerCreationInfo> layerCreationInfos = new List <LayerCreationInfo>();
for (int i = 1; i < layersNeuronsStr.Length; i++)
{
int layerIdx = i - 1;
LayerCreationInfo lci = new LayerCreationInfo();
lci.HowManyNeuronsPerLayer = int.Parse(layersNeuronsStr[i]);
lci.LayerNo = layerIdx;
lci.PreviousLayerNeuronsCount = layerIdx == 0 ? 0 : layerCreationInfos[layerIdx - 1].HowManyNeuronsPerLayer;
int LayerActivationFunctionInt = int.Parse(layersActivationFunctionStr[i]);
lci.LayerActivationFunction = GetActivationFunctionById(LayerActivationFunctionInt);
layerCreationInfos.Add(lci);
}
Topology topology = new Topology(layerCreationInfos, isNetworkUsingBias, minWeight, maxWeight);
neuralNetwork = new NeuralNetwork(topology, minWeight, maxWeight, methodOfLearning, name);
// provide saved neurons weights:
for (int layerNo = 0; layerNo < neuralNetwork.Topology.Layers.Count; layerNo++)
{
Layer layer = neuralNetwork.Topology.Layers[layerNo];
for (int neuronNo = 0; neuronNo < layer.Neurons.Count; neuronNo++)
{
String[] inputsString = sr.ReadLine().Split(';');
for (int inputNo = 0; inputNo < layer.Neurons[neuronNo].Inputs.Count; inputNo++)
{
layer.Neurons[neuronNo].Inputs[inputNo].Weight = double.Parse(inputsString[inputNo + 1]);
}
}
}
neuralNetwork.PropagateValuesForward(); // maybe not needed now, but for order it's good to get network in proper state.
return(neuralNetwork);
}
catch (Exception ex)
{
Console.WriteLine("Error while loading network: " + ex.Message);
return(null);
}
}