public void TestRPROPFolded()
{
IMLDataSet trainingData = XOR.CreateNoisyXORDataSet(10);
BasicNetwork network = NetworkUtil.CreateXORNetworkUntrained();
var folded = new FoldedDataSet(trainingData);
IMLTrain train = new ResilientPropagation(network, folded);
var trainFolded = new CrossValidationKFold(train, 4);
EncogUtility.TrainToError(trainFolded, 0.2);
XOR.VerifyXOR((IMLRegression)trainFolded.Method, 0.2);
}
/// <summary>
/// Create a trainer, use cross validation if enabled.
/// </summary>
/// <param name="method">The method to use.</param>
/// <param name="trainingSet">The training set to use.</param>
/// <returns>The trainer.</returns>
private IMLTrain CreateTrainer(IMLMethod method,
IMLDataSet trainingSet)
{
var factory = new MLTrainFactory();
String type = Prop.GetPropertyString(
ScriptProperties.MlTrainType);
String args = Prop.GetPropertyString(
ScriptProperties.MlTrainArguments);
EncogLogging.Log(EncogLogging.LevelDebug, "training type:" + type);
EncogLogging.Log(EncogLogging.LevelDebug, "training args:" + args);
IMLTrain train = factory.Create(method, trainingSet, type, args);
if (_kfold > 0)
{
train = new CrossValidationKFold(train, _kfold);
}
return(train);
}
public NeuralNetworkModel(VersatileMLDataSet dataset)
{
dataset.NormHelper.NormStrategy = new BasicNormalizationStrategy(0, 1, 0, 1);
dataset.Normalize();
var inputs = dataset.NormHelper.InputColumns.Count;
var outputs = dataset.NormHelper.OutputColumns.Count;
var hiddens = (inputs + outputs) * 1.5;
var method = (BasicNetwork) new MLMethodFactory().Create(
MLMethodFactory.TypeFeedforward,
$"?:B->SIGMOID->{hiddens}:B->SIGMOID->?",
inputs,
outputs);
var folds = new FoldedDataSet(dataset);
folds.Fold(5);
var propTrainer = new ResilientPropagation(method, folds);
_kfoldTrainer = new CrossValidationKFold(propTrainer, 5);
}
