MultiOutputGaussianProcess<T>
Models & Types · AiDotNet.GaussianProcesses
A Gaussian Process model that can predict multiple output values simultaneously.
Inheritance: ModelBase<T, Matrix<T>, Vector<T>> → GaussianProcessBase<T> → MultiOutputGaussianProcess<T>
Implements: IGeneratedParameterRegistrar<T>
For Beginners
A Gaussian Process is a flexible machine learning method that can learn patterns from data and provide uncertainty estimates with its predictions. Think of it as drawing a smooth curve through your data points, but also showing how confident it is about different parts of that curve.
This “Multi-Output” version can predict multiple related values at once. For example, if you’re predicting the temperature and humidity for weather forecasting, this model can learn how these outputs relate to each other and make better predictions by considering them together.
Unlike simpler models that just give you a single prediction, Gaussian Processes also tell you how confident they are about each prediction (the “variance” or “uncertainty”). This is especially useful when making decisions based on predictions where knowing the confidence level is important.
Constructors
| Constructor | Summary |
|---|---|
MultiOutputGaussianProcess(IKernelFunction<T> kernel, double? noiseVariance, int hyperparameterOptimizationSteps) |
Creates a new instance of the MultiOutputGaussianProcess with the specified kernel function. |
Methods
| Method | Summary |
|---|---|
ValueTuple<T, T> Predict(Vector<T> x) |
Predicts the mean and variance of a single-output model at x. |
ValueTuple<Vector<T>, Matrix<T>> PredictMultiOutput(Vector<T> x) |
Makes predictions for a new input point, returning both the predicted means and the covariance matrix. |
void Fit(Matrix<T> X, Vector<T> y) |
Fits the model to a single output - the IFullModel entry point, routed through FitMultiOutput. |
void FitMultiOutput(Matrix<T> X, Matrix<T> Y) |
Trains the Gaussian Process model on the provided multi-output training data. |
void UpdateKernel(IKernelFunction<T> kernel) |
Updates the kernel function used by the Gaussian Process and retrains the model if data is available. |