IMeanFunction<T>

Interfaces · AiDotNet.GaussianProcesses

Interface for mean functions in Gaussian Processes.

Known implementations: ConstantMean<T>, LinearMean<T>, PolynomialMean<T>, ZeroMean<T>

For Beginners

A mean function defines the “expected” value of the GP at any point before we observe any data. It represents our prior belief about the function’s behavior.

Common choices:

  • Zero mean: We expect the function to hover around zero
  • Constant mean: We expect the function to hover around some constant value
  • Linear mean: We expect a linear trend in the data

The GP then models deviations from this mean using the kernel function.

Methods

Method Summary
T Evaluate(Vector<T> x) Computes the mean values at multiple input points.
Vector<T> Evaluate(Matrix<T> X) Computes the mean values at multiple input points.