rational_quadratic

luas.kernels.rational_quadratic(x: Array, y: Array, L: Any, alpha: Any) → Array

Rational quadratic kernel function, used with luas.kernels.evaluate_kernel to build covariance matrices.

\[k(x, y) = \Bigg(1 + \frac{|x - y|^2}{2 \alpha L^2}\Bigg)^{-\alpha}\]

Parameters:

• x (JAXArray) – Input vector 1

• y (JAXArray) – Input vector 2

• L (Scalar) – Length scale

• alpha (Scalar) – Scale mixture parameter

Returns:

Covariance between two input vectors

Return type:

Scalar