MultiOutputRegressor<T>

Models & Types · AiDotNet.Regression.MultiOutput

Wraps a single-output regression model into a MULTI-OUTPUT one: given an n×H target, it trains H independent “head” models (one per target column, the direct-multi-horizon strategy) and predicts an n×H matrix. This is the regression analogue of the multi-label classifier chain, and the substrate the shape-polymorphic facade routes to when the training target has more than one column (e.g. a multi-horizon forecast vector).

Implements: IFeatureAware, IFullModel<T, Matrix<T>, Matrix<T>>

How It Works

For Beginners: some models (linear/forest/boosting) only predict a single number. To forecast several horizons at once (1, 5, 10, 20 bars ahead), this trains one copy per horizon behind a single object, so the caller trains once with an n×H target and gets an n×H prediction back — no per-horizon plumbing. Deep sequence models that emit a horizon vector natively don’t need this; they output n×H directly.

Constructors

Constructor Summary
MultiOutputRegressor(Func<IFullModel<T, Matrix<T>, Vector<T>>> baseFactory)

Properties

Property Summary
ILossFunction<T> DefaultLossFunction { get; }
bool SupportsParameterInitialization { get; }
int OutputCount { get; } Number of output columns (horizons) the last Train produced.
long ParameterCount { get; }

Methods

Method Summary
Dictionary<string, T> GetFeatureImportance()
IEnumerable<int> GetActiveFeatureIndices()
IFullModel<T, Matrix<T>, Matrix<T>> Clone()
IFullModel<T, Matrix<T>, Matrix<T>> DeepCopy()
IFullModel<T, Matrix<T>, Matrix<T>> WithParameters(Vector<T> parameters)
Matrix<T> Predict(Matrix<T> input)
ModelMetadata<T> GetModelMetadata()
Vector<T> GetParameters()
Vector<T> SanitizeParameters(Vector<T> parameters)
bool IsFeatureUsed(int featureIndex)
byte[] Serialize()
void Deserialize(byte[] data)
void Dispose()
void LoadModel(string filePath)
void LoadState(Stream stream)
void SaveModel(string filePath)
void SaveState(Stream stream)
void SetActiveFeatureIndices(IEnumerable<int> featureIndices)
void SetParameters(Vector<T> parameters)
void Train(Matrix<T> input, Matrix<T> expectedOutput)