/// <summary>
/// Creates the instance and initializes it from given xml element.
/// This is the preferred way to instantiate ReadoutLayer settings.
/// </summary>
/// <param name="elem">
/// Xml data containing ReadoutLayer settings.
/// Content of xml element is always validated against the xml schema.
/// </param>
public ReadoutLayerSettings(XElement elem)
{
//Validation
ElemValidator validator = new ElemValidator();
Assembly assemblyRCNet = Assembly.GetExecutingAssembly();
validator.AddXsdFromResources(assemblyRCNet, "RCNet.Neural.Network.SM.ReadoutLayerSettings.xsd");
validator.AddXsdFromResources(assemblyRCNet, "RCNet.RCNetTypes.xsd");
XElement readoutLayerSettingsElem = validator.Validate(elem, "rootElem");
//Parsing
TestDataRatio = double.Parse(readoutLayerSettingsElem.Attribute("testDataRatio").Value, CultureInfo.InvariantCulture);
NumOfFolds = readoutLayerSettingsElem.Attribute("folds").Value == "Auto" ? 0 : int.Parse(readoutLayerSettingsElem.Attribute("folds").Value);
//Readout unit
XElement readoutUnitElem = readoutLayerSettingsElem.Descendants("readoutUnit").First();
ReadoutUnitCfg = new ReadoutUnitSettings(readoutUnitElem);
//Output fields
XElement outputFieldsElem = readoutLayerSettingsElem.Descendants("outputFields").First();
OutputFieldNameCollection = new List <string>();
foreach (XElement outputFieldElem in outputFieldsElem.Descendants("field"))
{
OutputFieldNameCollection.Add(outputFieldElem.Attribute("name").Value);
}
return;
}
/// <summary>
/// Creates the readout layer configuration.
/// </summary>
/// <param name="foldDataRatio">Specifies what part of available data to be used as the fold data.</param>
/// <param name="numOfAttempts">Number of regression attempts. Each readout network will try to learn numOfAttempts times.</param>
/// <param name="numOfEpochs">Number of training epochs within an attempt.</param>
ReadoutLayerSettings CreateReadoutLayerCfg(double foldDataRatio, int numOfAttempts, int numOfEpochs)
{
//For each output field we will use prediction of two networks
//First network having only Identity output neuron and associated the resilient back propagation trainer
FeedForwardNetworkSettings ffNet1Cfg = new FeedForwardNetworkSettings(new AFAnalogIdentitySettings(),
null,
new RPropTrainerSettings(numOfAttempts, numOfEpochs)
);
//Second network having Identity output neuron, hidden layer consisting of 5 LeakyReLU neurons
//and associated the resilient back propagation trainer
HiddenLayerSettings hiddenLayerCfg = new HiddenLayerSettings(5, new AFAnalogLeakyReLUSettings());
FeedForwardNetworkSettings ffNet2Cfg = new FeedForwardNetworkSettings(new AFAnalogIdentitySettings(),
new HiddenLayersSettings(hiddenLayerCfg),
new RPropTrainerSettings(numOfAttempts, numOfEpochs)
);
//Create the cluster chain configuration for the forecast and the default configuration for the forecast task.
CrossvalidationSettings crossvalidationCfg = new CrossvalidationSettings(foldDataRatio);
TNRNetClusterRealNetworksSettings networksCfg = new TNRNetClusterRealNetworksSettings(ffNet1Cfg, ffNet2Cfg);
TNRNetClusterRealSettings realClusterCfg = new TNRNetClusterRealSettings(networksCfg, new TNRNetClusterRealWeightsSettings());
TNRNetClusterChainRealSettings realClusterChainCfg = new TNRNetClusterChainRealSettings(crossvalidationCfg, new TNRNetClustersRealSettings(realClusterCfg));
TaskDefaultsSettings taskDefaultsCfg = new TaskDefaultsSettings(null, realClusterChainCfg);
//Create readout unit configurations. We will forecast next High and Low prices.
ReadoutUnitSettings highReadoutUnitCfg = new ReadoutUnitSettings("High", new ForecastTaskSettings());
ReadoutUnitSettings lowReadoutUnitCfg = new ReadoutUnitSettings("Low", new ForecastTaskSettings());
//Create readout layer configuration
ReadoutLayerSettings readoutLayerCfg = new ReadoutLayerSettings(taskDefaultsCfg,
new ReadoutUnitsSettings(highReadoutUnitCfg,
lowReadoutUnitCfg
),
null
);
return(readoutLayerCfg);
}
/// <summary>
/// Creates readout layer configuration
/// </summary>
/// <param name="testDataRatio">Specifies what part of available data to be used as test data</param>
/// <param name="numOfAttempts">Number of regression attempts. Each readout network will try to learn numOfAttempts times</param>
/// <param name="numOfEpochs">Number of training epochs within an attempt</param>
ReadoutLayerSettings CreateReadoutLayerCfg(double testDataRatio, int numOfAttempts, int numOfEpochs)
{
//For each output field we will use prediction of two networks
//First network having only Identity output neuron and associated the resilient back propagation trainer
FeedForwardNetworkSettings ffNet1Cfg = new FeedForwardNetworkSettings(new IdentitySettings(),
null,
new RPropTrainerSettings(numOfAttempts, numOfEpochs)
);
//Second network having Identity output neuron, hidden layer consisting of 5 LeakyReLU neurons
//and associated the resilient back propagation trainer
HiddenLayerSettings hiddenLayerCfg = new HiddenLayerSettings(5, new LeakyReLUSettings());
FeedForwardNetworkSettings ffNet2Cfg = new FeedForwardNetworkSettings(new IdentitySettings(),
new HiddenLayersSettings(hiddenLayerCfg),
new RPropTrainerSettings(numOfAttempts, numOfEpochs)
);
//Create default networks configuration for forecasting
DefaultNetworksSettings defaultNetworksCfg = new DefaultNetworksSettings(null, new ForecastNetworksSettings(ffNet1Cfg, ffNet2Cfg));
//Create readout units. We will forecast next High and Low prices. Both fields are real numbers.
ReadoutUnitSettings highReadoutUnitCfg = new ReadoutUnitSettings("High", new ForecastTaskSettings(new RealFeatureFilterSettings()));
ReadoutUnitSettings lowReadoutUnitCfg = new ReadoutUnitSettings("Low", new ForecastTaskSettings(new RealFeatureFilterSettings()));
//Create readout layer configuration
ReadoutLayerSettings readoutLayerCfg = new ReadoutLayerSettings(new ReadoutUnitsSettings(highReadoutUnitCfg,
lowReadoutUnitCfg
),
testDataRatio,
ReadoutLayerSettings.AutoFolds,
ReadoutLayerSettings.DefaultRepetitions,
defaultNetworksCfg
);
return(readoutLayerCfg);
}
//Constructors
/// <summary>
/// Creates an uninitialized instance.
/// </summary>
public ReadoutLayerSettings()
{
//Default settings
TestDataRatio = 0;
NumOfFolds = 0;
ReadoutUnitCfg = new ReadoutUnitSettings();
OutputFieldNameCollection = new List <string>();
return;
}
//Constructors
/// <summary>
/// Creates an uninitialized instance.
/// </summary>
/// <param name="unitCfg">The configuration of the readout unit.</param>
/// <param name="unitIndex">An index of the readout unit within the readout layer.</param>
/// <param name="memberIndex">An index within the "One Takes All" group.</param>
public MemberErrorStat(ReadoutUnitSettings unitCfg, int unitIndex, int memberIndex)
{
UnitCfg = (ReadoutUnitSettings)unitCfg.DeepClone();
UnitIndex = unitIndex;
MemberIndex = memberIndex;
ErrStat = new BasicStat();
NumOfCorrectButBellowBorderSelections = 0;
NumOfOverbeatedAboveBorderRawProbabilities = 0;
return;
}
//Constructor
/// <summary>
/// Creates an unitialized instance
/// </summary>
/// <param name="index">The zero-based index of the readout unit.</param>
/// <param name="readoutUnitCfg">The configuration of the readout unit.</param>
public ReadoutUnitErrorStat(int index, ReadoutUnitSettings readoutUnitCfg)
{
Name = readoutUnitCfg.Name;
Index = index;
Task = readoutUnitCfg.TaskCfg.Type;
ErrorStat = new BasicStat();
if (Task == ReadoutUnit.TaskType.Classification)
{
BinErrorStat = new BinErrStat(0.5d);
}
return;
}
//Methods
/// <summary>
/// See the base.
/// </summary>
public override bool Equals(object obj)
{
if (obj == null)
{
return(false);
}
ReadoutUnitSettings cmpSettings = obj as ReadoutUnitSettings;
if (NetType != cmpSettings.NetType ||
!Equals(NetSettings, cmpSettings.NetSettings) ||
RegressionAttempts != cmpSettings.RegressionAttempts ||
RegressionAttemptEpochs != cmpSettings.RegressionAttemptEpochs
)
{
return(false);
}
return(true);
}
/// <summary>
/// Copy constructor
/// </summary>
/// <param name="source">Source instance</param>
public ReadoutUnitSettings(ReadoutUnitSettings source)
{
NetType = source.NetType;
NetSettings = null;
if (source.NetSettings != null)
{
if (source.NetSettings.GetType() == typeof(FeedForwardNetworkSettings))
{
NetSettings = ((FeedForwardNetworkSettings)source.NetSettings).DeepClone();
}
else
{
NetSettings = ((ParallelPerceptronSettings)source.NetSettings).DeepClone();
}
}
RegressionAttempts = source.RegressionAttempts;
RegressionAttemptEpochs = source.RegressionAttemptEpochs;
return;
}
//Methods
/// <summary>
/// Checks consistency
/// </summary>
protected override void Check()
{
if (MapperCfg != null && NeuralPreprocessorCfg == null)
{
throw new ArgumentException($"Mapper can not be specified when neural preprocessor is not defined.", "MapperCfg");
}
if (MapperCfg != null)
{
foreach (ReadoutUnitMapSettings map in MapperCfg.MapCfgCollection)
{
ReadoutUnitSettings rus = ReadoutLayerCfg.ReadoutUnitsCfg.GetReadoutunitCfg(map.ReadoutUnitName);
//Pools
if (map.AllowedPoolsCfg != null)
{
foreach (AllowedPoolSettings aps in map.AllowedPoolsCfg.AllowedPoolCfgCollection)
{
ReservoirInstanceSettings ris = NeuralPreprocessorCfg.ReservoirInstancesCfg.GetReservoirInstanceCfg(aps.ReservoirInstanceName);
ReservoirStructureSettings rss = NeuralPreprocessorCfg.ReservoirStructuresCfg.GetReservoirStructureCfg(ris.StructureCfgName);
rss.PoolsCfg.GetPoolID(aps.PoolName);
}
}
//Input fields
if (map.AllowedInputFieldsCfg != null)
{
string[] routedFieldNames = NeuralPreprocessorCfg.InputEncoderCfg.GetRoutedFieldNames().ToArray();
foreach (AllowedInputFieldSettings aifs in map.AllowedInputFieldsCfg.AllowedInputFieldCfgCollection)
{
if (Array.IndexOf(routedFieldNames, aifs.Name) == -1)
{
throw new ArgumentException($"Specified input field {aifs.Name} to be allowed for readout unit {map.ReadoutUnitName} is not among fields routed to readout layer.", "MapperCfg");
}
}
}
}
}
return;
}