/// <summary>
/// Create an annealing trainer.
/// </summary>
///
/// <param name="method">The method to use.</param>
/// <param name="training">The training data to use.</param>
/// <param name="argsStr">The arguments to use.</param>
/// <returns>The newly created trainer.</returns>
public IMLTrain Create(IMLMethod method,
IMLDataSet training, String argsStr)
{
if (!(method is BasicNetwork))
{
throw new TrainingError(
"Invalid method type, requires BasicNetwork");
}
ICalculateScore score = new TrainingSetScore(training);
IDictionary <String, String> args = ArchitectureParse.ParseParams(argsStr);
var holder = new ParamsHolder(args);
double startTemp = holder.GetDouble(
MLTrainFactory.PropertyTemperatureStart, false, 10);
double stopTemp = holder.GetDouble(
MLTrainFactory.PropertyTemperatureStop, false, 2);
int cycles = holder.GetInt(MLTrainFactory.Cycles, false, 100);
IMLTrain train = new NeuralSimulatedAnnealing(
(BasicNetwork)method, score, startTemp, stopTemp, cycles);
return(train);
}
private double TrainNetwork(String what, BasicNetwork network, IMLDataSet trainingSet, string Method)
{
// train the neural network
ICalculateScore score = new TrainingSetScore(trainingSet);
IMLTrain trainAlt = new NeuralSimulatedAnnealing(network, score, 10, 2, 100);
IMLTrain trainMain;
if (Method.Equals("Leven"))
{
Console.WriteLine("Using LevenbergMarquardtTraining");
trainMain = new LevenbergMarquardtTraining(network, trainingSet);
}
else
{
trainMain = new Backpropagation(network, trainingSet);
}
var stop = new StopTrainingStrategy();
trainMain.AddStrategy(new Greedy());
trainMain.AddStrategy(new HybridStrategy(trainAlt));
trainMain.AddStrategy(stop);
int epoch = 0;
while (!stop.ShouldStop())
{
trainMain.Iteration();
app.WriteLine("Training " + what + ", Epoch #" + epoch + " Error:" + trainMain.Error);
epoch++;
}
return(trainMain.Error);
}
public override double Train(Data info, float lr, float mom)
{
IMLDataSet data = new BasicMLDataSet(info.InputData, info.OutputData);
//Train network on data set, parameters (Network, dataset, learning rate, momentum).
ICalculateScore score = new TrainingSetScore(data);
IMLTrain trainAlt = new NeuralSimulatedAnnealing(EncogNetwork, score, 10, 2, 1000);
IMLTrain learner;
learner = new LevenbergMarquardtTraining(EncogNetwork, data);
var stop = new StopTrainingStrategy();
learner.AddStrategy(new Greedy());
learner.AddStrategy(new HybridStrategy(trainAlt));
//Train network on data set.
double lastError = double.PositiveInfinity;
do
{
if (learner.Error != 0)
{
lastError = learner.Error;
}
learner.Iteration();
} while (lastError - learner.Error > 0.0000001);
return(learner.Error);
}
public static double TrainNetworks(BasicNetwork network, IMLDataSet minis)
{
// train the neural network
ICalculateScore score = new TrainingSetScore(minis);
IMLTrain trainAlt = new NeuralSimulatedAnnealing(network, score, 10, 2, 100);
IMLTrain trainMain = new Backpropagation(network, minis, 0.0001, 0.01);
StopTrainingStrategy stop = new StopTrainingStrategy(0.0001, 200);
trainMain.AddStrategy(new Greedy());
trainMain.AddStrategy(new HybridStrategy(trainAlt));
trainMain.AddStrategy(stop);
var sw = new Stopwatch();
sw.Start();
while (!stop.ShouldStop())
{
trainMain.Iteration();
Console.WriteLine(@"Iteration #:" + trainMain.IterationNumber + @" Error:" + trainMain.Error + @" Genetic Iteration:" + trainAlt.IterationNumber);
}
sw.Stop();
return(trainMain.Error);
}
/// <summary>
/// Create an annealing trainer.
/// </summary>
///
/// <param name="method">The method to use.</param>
/// <param name="training">The training data to use.</param>
/// <param name="argsStr">The arguments to use.</param>
/// <returns>The newly created trainer.</returns>
public IMLTrain Create(IMLMethod method,
IMLDataSet training, String argsStr)
{
if (!(method is BasicNetwork))
{
throw new TrainingError(
"Invalid method type, requires BasicNetwork");
}
ICalculateScore score = new TrainingSetScore(training);
IDictionary <String, String> args = ArchitectureParse.ParseParams(argsStr);
var holder = new ParamsHolder(args);
int populationSize = holder.GetInt(
MLTrainFactory.PropertyPopulationSize, false, 5000);
double mutation = holder.GetDouble(
MLTrainFactory.PropertyMutation, false, 0.1d);
double mate = holder.GetDouble(MLTrainFactory.PropertyMate,
false, 0.25d);
IMLTrain train = new NeuralGeneticAlgorithm((BasicNetwork)method,
new RangeRandomizer(-1, 1), score, populationSize, mutation,
mate);
return(train);
}
/// <summary>
/// Create an annealing trainer.
/// </summary>
/// <param name="method">The method to use.</param>
/// <param name="training">The training data to use.</param>
/// <param name="argsStr">The arguments to use.</param>
/// <returns>The newly created trainer.</returns>
public IMLTrain Create(IMLMethod method,
IMLDataSet training, String argsStr)
{
if (!(method is BasicNetwork))
{
throw new TrainingError(
"Invalid method type, requires BasicNetwork");
}
ICalculateScore score = new TrainingSetScore(training);
IDictionary <String, String> args = ArchitectureParse.ParseParams(argsStr);
var holder = new ParamsHolder(args);
int populationSize = holder.GetInt(
MLTrainFactory.PropertyPopulationSize, false, 5000);
IMLTrain train = new MLMethodGeneticAlgorithm(() =>
{
IMLMethod result = (IMLMethod)ObjectCloner.DeepCopy(method);
((IMLResettable)result).Reset();
return(result);
}, score, populationSize);
return(train);
}
private double TrainNetwork(String what, BasicNetwork network, IMLDataSet trainingSet)
{
// train the neural network
ICalculateScore score = new TrainingSetScore(trainingSet);
IMLTrain trainAlt = new NeuralSimulatedAnnealing(
network, score, 10, 2, 100);
IMLTrain trainMain = new Backpropagation(network, trainingSet, 0.00001, 0.0);
var stop = new StopTrainingStrategy();
trainMain.AddStrategy(new Greedy());
trainMain.AddStrategy(new HybridStrategy(trainAlt));
trainMain.AddStrategy(stop);
int epoch = 0;
while (!stop.ShouldStop())
{
trainMain.Iteration();
app.WriteLine("Training " + what + ", Epoch #" + epoch + " Error:" + trainMain.Error);
epoch++;
}
return(trainMain.Error);
}
static void Main(string[] args)
{
// this form of ANN uses genetic algorithm to produce
// hidden layer of neurons
// A NEAT network starts with only an
// input layer and output layer. The rest is evolved as the training progresses.
// Connections inside of a NEAT neural network can be feedforward, recurrent,
// or self - connected.All of these connection types will be tried by NEAT as it
// attempts to evolve a neural network capable of the given task.
IMLDataSet trainingSet = new BasicMLDataSet(XORInput, XORIdeal);
NEATPopulation pop = new NEATPopulation(2, 1, 1000);
pop.Reset();
pop.InitialConnectionDensity = 1.0; // not required, but speeds processing.
ICalculateScore score = new TrainingSetScore(trainingSet);
// train the neural network
TrainEA train = NEATUtil.ConstructNEATTrainer(pop, score);
EncogUtility.TrainToError(train, 0.01);
NEATNetwork network = (NEATNetwork)train.CODEC.Decode(train.BestGenome);
// TODO no persistance? no means to peek structure?
// test the neural network
Console.WriteLine(@"Neural Network Results:");
EncogUtility.Evaluate(network, trainingSet);
}
private void trainNetworkBackprop()
{
// IMLTrain train = new Backpropagation(this.network, this.input,this.ideal, 0.000001, 0.1);
IMLDataSet aset = new BasicMLDataSet(input, ideal);
int epoch = 1;
// train the neural network
ICalculateScore score = new TrainingSetScore(aset);
IMLTrain trainAlt = new NeuralSimulatedAnnealing(network, score, 10, 2, 100);
IMLTrain trainMain = new Backpropagation(network, aset, 0.001, 0.0);
StopTrainingStrategy stop = new StopTrainingStrategy();
var pop = new NEATPopulation(INPUT_SIZE, OUTPUT_SIZE, 1000);
// train the neural network
var step = new ActivationStep();
step.Center = 0.5;
pop.OutputActivationFunction = step;
var train = new NEATTraining(score, pop);
trainMain.AddStrategy(new Greedy());
trainMain.AddStrategy(new HybridStrategy(trainAlt));
trainMain.AddStrategy(stop);
trainMain.AddStrategy(new HybridStrategy(train));
network.ClearContext();
while (!stop.ShouldStop())
{
trainMain.Iteration();
train.Iteration();
Console.WriteLine(@"Training " + @"Epoch #" + epoch + @" Error:" + trainMain.Error + @" Genetic iteration:" + trainAlt.IterationNumber + @"neat iteration:" + train.IterationNumber);
epoch++;
}
}
/// <summary>
/// Create an NEAT GA trainer.
/// </summary>
/// <param name="method">The method to use.</param>
/// <param name="training">The training data to use.</param>
/// <param name="argsStr">The arguments to use.</param>
/// <returns>The newly created trainer.</returns>
public IMLTrain Create(IMLMethod method,
IMLDataSet training, String argsStr)
{
ICalculateScore score = new TrainingSetScore(training);
TrainEA train = NEATUtil.ConstructNEATTrainer((NEATPopulation)method, score);
return(train);
}
private void Validate(NEATPopulation pop)
{
Assert.AreEqual(10, pop.PopulationSize);
Assert.AreEqual(0.2, pop.SurvivalRate);
// see if the population can actually be used to train
IMLDataSet trainingSet = new BasicMLDataSet(XOR.XORInput, XOR.XORIdeal);
ICalculateScore score = new TrainingSetScore(trainingSet);
IEvolutionaryAlgorithm train = NEATUtil.ConstructNEATTrainer(pop, score);
train.Iteration();
}
private NEATPopulation Generate()
{
IMLDataSet trainingSet = new BasicMLDataSet(XOR.XORInput, XOR.XORIdeal);
ICalculateScore score = new TrainingSetScore(trainingSet);
// train the neural network
ActivationStep step = new ActivationStep();
step.Center = 0.5;
IEvolutionaryAlgorithm train = NEATUtil.ConstructNEATTrainer(
score, 2, 1, 10);
return((NEATPopulation)train.Population);
}
private NEATPopulation Generate()
{
IMLDataSet trainingSet = new BasicMLDataSet(XOR.XORInput, XOR.XORIdeal);
ICalculateScore score = new TrainingSetScore(trainingSet);
// train the neural network
ActivationStep step = new ActivationStep();
step.Center = 0.5;
NEATTraining train = new NEATTraining(
score, 2, 1, 10);
return((NEATPopulation)train.Population);
}
/// <summary>
/// Builds and trains a neat network.
/// </summary>
/// <param name="aset">The IMLDataset.</param>
/// <param name="inputcounts">The inputcounts.</param>
/// <param name="outputcounts">The outputcounts.</param>
/// <param name="populationsize">The populationsize.</param>
/// <param name="ToErrorTraining">To error rate you want to train too.</param>
/// <returns>a trained netnetwork.</returns>
public static NEATNetwork BuildTrainNeatNetwork(IMLDataSet aset, int inputcounts, int outputcounts, int populationsize, double ToErrorTraining)
{
NEATPopulation pop = new NEATPopulation(inputcounts, outputcounts, populationsize);
ICalculateScore score = new TrainingSetScore(aset);
// train the neural network
ActivationStep step = new ActivationStep();
step.Center = 0.5;
pop.OutputActivationFunction = step;
NEATTraining train = new NEATTraining(score, pop);
EncogUtility.TrainToError(train, ToErrorTraining);
NEATNetwork network = (NEATNetwork)train.Method;
return(network);
}
private void Validate(NEATPopulation pop)
{
Assert.AreEqual(0.3, pop.OldAgePenalty);
Assert.AreEqual(50, pop.OldAgeThreshold);
Assert.AreEqual(10, pop.PopulationSize);
Assert.AreEqual(0.2, pop.SurvivalRate);
Assert.AreEqual(10, pop.YoungBonusAgeThreshold);
Assert.AreEqual(0.3, pop.YoungScoreBonus);
// see if the population can actually be used to train
IMLDataSet trainingSet = new BasicMLDataSet(XOR.XORInput, XOR.XORIdeal);
ICalculateScore score = new TrainingSetScore(trainingSet);
NEATTraining train = new NEATTraining(score, pop);
train.Iteration();
}
/// <summary>
/// Create a PSO trainer.
/// </summary>
/// <param name="method">The method to use.</param>
/// <param name="training">The training data to use.</param>
/// <param name="argsStr">The arguments to use.</param>
/// <returns>The newly created trainer.</returns>
public IMLTrain Create(IMLMethod method,
IMLDataSet training, String argsStr)
{
IDictionary <String, String> args = ArchitectureParse.ParseParams(argsStr);
ParamsHolder holder = new ParamsHolder(args);
int particles = holder.GetInt(
MLTrainFactory.PropertyParticles, false, 20);
ICalculateScore score = new TrainingSetScore(training);
IRandomizer randomizer = new NguyenWidrowRandomizer();
IMLTrain train = new NeuralPSO((BasicNetwork)method, randomizer, score, particles);
return(train);
}
/// <summary>
/// Create an EPL GA trainer.
/// </summary>
/// <param name="method">The method to use.</param>
/// <param name="training">The training data to use.</param>
/// <param name="argsStr">The arguments to use.</param>
/// <returns>The newly created trainer.</returns>
public IMLTrain Create(IMLMethod method,
IMLDataSet training, String argsStr)
{
PrgPopulation pop = (PrgPopulation)method;
ICalculateScore score = new TrainingSetScore(training);
TrainEA train = new TrainEA(pop, score);
train.Rules.AddRewriteRule(new RewriteConstants());
train.Rules.AddRewriteRule(new RewriteAlgebraic());
train.CODEC = new PrgCODEC();
train.AddOperation(0.8, new SubtreeCrossover());
train.AddOperation(0.1, new SubtreeMutation(pop.Context, 4));
train.AddOperation(0.1, new ConstMutation(pop.Context, 0.5, 1.0));
train.AddScoreAdjuster(new ComplexityAdjustedScore());
train.Speciation = new PrgSpeciation();
return(train);
}
/// <summary>
/// Program entry point.
/// </summary>
/// <param name="app">Holds arguments and other info.</param>
public void Execute(IExampleInterface app)
{
IMLDataSet trainingSet = new BasicMLDataSet(XORInput, XORIdeal);
BasicNetwork network = EncogUtility.SimpleFeedForward(2, 2, 0, 1, false);
ICalculateScore score = new TrainingSetScore(trainingSet);
IRandomizer randomizer = new NguyenWidrowRandomizer(-1, 1);
IMLTrain train = new NeuralPSO(network, randomizer, score, 20);
EncogUtility.TrainToError(train, 0.01);
network = (BasicNetwork)train.Method;
// test the neural network
Console.WriteLine("Neural Network Results:");
EncogUtility.Evaluate(network, trainingSet);
EncogFramework.Instance.Shutdown();
}
/// <summary>
/// The entry point for this example. If you would like to make this example
/// stand alone, then add to its own project and rename to Main.
/// </summary>
/// <param name="args">Not used.</param>
public static void ExampleMain(string[] args)
{
IMLDataSet trainingSet = new BasicMLDataSet(XORInput, XORIdeal);
var pop = new NEATPopulation(2, 1, 1000);
pop.Reset();
pop.InitialConnectionDensity = 1.0; // not required, but speeds processing.
ICalculateScore score = new TrainingSetScore(trainingSet);
// train the neural network
var train = NEATUtil.ConstructNEATTrainer(pop, score);
EncogUtility.TrainToError(train, 0.01);
var network = (NEATNetwork)train.CODEC.Decode(train.BestGenome);
// test the neural network
Console.WriteLine(@"Neural Network Results:");
EncogUtility.Evaluate(network, trainingSet);
}
public static double TrainNetwork(String what,
FreeformNetwork network, IMLDataSet trainingSet)
{
ICalculateScore score = new TrainingSetScore(trainingSet);
IMLTrain trainAlt = new NeuralSimulatedAnnealing(
network, score, 10, 2, 100);
IMLTrain trainMain = new FreeformBackPropagation(network, trainingSet, 0.00001, 0.0);
StopTrainingStrategy stop = new StopTrainingStrategy();
trainMain.AddStrategy(new Greedy());
trainMain.AddStrategy(new HybridStrategy(trainAlt));
trainMain.AddStrategy(stop);
EncogUtility.TrainToError(trainMain, 0.01);
return(trainMain.Error);
}
public void Train(BasicNetwork network, IMLDataSet training)
{
IMLTrain trainMain = new LevenbergMarquardtTraining(network, training);
// train the neural network
var stop = new StopTrainingStrategy();
var score = new TrainingSetScore(trainMain.Training);
var trainAlt = new NeuralSimulatedAnnealing(network, score, 10, 2, 100);
trainMain.AddStrategy(new HybridStrategy(trainAlt));
trainMain.AddStrategy(stop);
var epoch = 0;
while (!stop.ShouldStop() && trainMain.IterationNumber < 1500)
{
trainMain.Iteration();
Console.WriteLine("Training " + ", Epoch #" + epoch + " Error:" + trainMain.Error);
epoch++;
}
}
/// <summary>
/// Program entry point.
/// </summary>
/// <param name="app">Holds arguments and other info.</param>
public void Execute(IExampleInterface app)
{
IMLDataSet trainingSet = new BasicMLDataSet(XORInput, XORIdeal);
var pop = new NEATPopulation(2, 1, 1000);
ICalculateScore score = new TrainingSetScore(trainingSet);
// train the neural network
var step = new ActivationStep();
step.Center = 0.5;
pop.OutputActivationFunction = step;
var train = new NEATTraining(score, pop);
EncogUtility.TrainToError(train, 0.01);
var network = (NEATNetwork)train.Method;
network.ClearContext();
// test the neural network
Console.WriteLine(@"Neural Network Results:");
EncogUtility.Evaluate(network, trainingSet);
}
public static double TrainNetworks(BasicNetwork network, IMLDataSet minis)
{
Backpropagation trainMain = new Backpropagation(network, minis, 0.0001, 0.6);
//set the number of threads below.
trainMain.ThreadCount = 0;
// train the neural network
ICalculateScore score = new TrainingSetScore(minis);
IMLTrain trainAlt = new NeuralSimulatedAnnealing(network, score, 10, 2, 100);
// IMLTrain trainMain = new Backpropagation(network, minis, 0.0001, 0.01);
StopTrainingStrategy stop = new StopTrainingStrategy(0.0001, 200);
trainMain.AddStrategy(new Greedy());
trainMain.AddStrategy(new HybridStrategy(trainAlt));
trainMain.AddStrategy(stop);
//prune strategy not in GIT!...Removing it.
//PruneStrategy strategypruning = new PruneStrategy(0.91d, 0.001d, 10, network,minis, 0, 20);
//trainMain.AddStrategy(strategypruning);
EncogUtility.TrainConsole(trainMain, network, minis, 15.2);
var sw = new Stopwatch();
sw.Start();
while (!stop.ShouldStop())
{
trainMain.Iteration();
Console.WriteLine(@"Iteration #:" + trainMain.IterationNumber + @" Error:" + trainMain.Error + @" Genetic Iteration:" + trainAlt.IterationNumber);
}
sw.Stop();
Console.WriteLine(@"Total elapsed time in seconds:" + TimeSpan.FromMilliseconds(sw.ElapsedMilliseconds).Seconds);
return(trainMain.Error);
}
/// <summary>
/// Trains the network
/// </summary>
public virtual void trainNetwork()
{
INeuralDataSet trainingSet = new BasicNeuralDataSet(networkInput, networkIdealOutput);
//ITrain trainBackProp = new Backpropagation(network, trainingSet, BACKPROP_LEARN_RATE, BACKPROP_MOMENTUM);
ITrain trainBackProp = new ScaledConjugateGradient(network, trainingSet);
double error = Double.MaxValue;
double lastError = Double.MaxValue;
int epoch = 1;
int lastAnneal = 0;
int errorExit = 0;
double errorOnLastAnnealStart = double.MaxValue;
double sameErrorOnLastAnnealStartCount = 0;
double currentAnnealInterval = MIN_BACKPROP_ITERATIONS_ANNEAL_START;
double annealStartError = 0;
do
{
trainBackProp.Iteration();
error = trainBackProp.Error;
if (lastError - error < MAX_RMS_ITERATION_NETWORK_ERROR)
errorExit++;
else
errorExit = 0;
Console.WriteLine("Iteration(SC) #{0} Error: {1}", epoch, error.ToString("0.00000000"));
if (error > ANNEAL_MIN_ERROR)
{
if ((lastAnneal > currentAnnealInterval) && (lastError - error < MAX_ANNEAL_START_ERROR))
{
if (error == errorOnLastAnnealStart)
sameErrorOnLastAnnealStartCount++;
else if (error < errorOnLastAnnealStart)
{
sameErrorOnLastAnnealStartCount = 0;
errorOnLastAnnealStart = error;
}
ICalculateScore score = new TrainingSetScore(trainingSet);
NeuralSimulatedAnnealing trainAnneal = new NeuralSimulatedAnnealing(network, score, ANNEAL_STARTTEMP, ANNEAL_ENDTEMP, ANNEAL_ITERATIONS);
for (int i = 1; i <= ANNEAL_ATTEMPTS; i++)
{
trainAnneal.Iteration();
if (i == 1)
annealStartError = trainAnneal.Error;
Console.WriteLine("Iteration(Anneal) #{0}-{1} Error: {2}", epoch, i, trainAnneal.Error.ToString("0.00000000"));
//WebLogging.AddLog("WinRatioNeural", WebLogging.LogCategory.WinRatioNeural, "Iteration(Anneal) #" + i + " Error: " + trainAnneal.Error.ToString("0.00000000"));
}
if (annealStartError == trainAnneal.Error)
{
if (currentAnnealInterval < 200)
{
currentAnnealInterval *= 1.5;
Console.WriteLine("Iteration(Anneal) # No improvment. Increasing anneal interval to " + currentAnnealInterval);
}
else
Console.WriteLine("Iteration(Anneal) # No improvment. Anneal interval at max.");
}
lastAnneal = 0;
trainBackProp = new ScaledConjugateGradient(network, trainingSet);
trainBackProp.Iteration();
error = trainBackProp.Error;
//saveNetwork(correctPredictions.ToString("##0.0")+ "_" + epoch.ToString() + "_nerualPokerAI_LA.nnDAT");
}
}
//Every 50 epochs we can test the network accuracy
//#if DEBUG
//if (epoch % 50 == 0)
//{
// //We want to switch to the testing set if we are not using all data for training
// if (TRAIN_DATA_PERCENT < 1.0) createTestingSets();
// Console.WriteLine(" Network accuracy is currently {0}%",getNetworkAccuracy());
// //Wait for 1 second so that we can read the output
// Thread.Sleep(1000);
// //Likewise we want to switch back before continuing
// if (TRAIN_DATA_PERCENT < 1.0) createTrainingSets();
//}
//#endif
lastError = trainBackProp.Error;
epoch++;
lastAnneal++;
//} while (error > MAX_RMS_TOTAL_NETWORK_ERROR && errorExit < 10 && epoch < MAX_ITERATIONS);
} while (trainBackProp.Error > MAX_RMS_TOTAL_NETWORK_ERROR && epoch < MAX_ITERATIONS && sameErrorOnLastAnnealStartCount < 2);
}
