Gaussian process (GP) regression has been widely used in supervised machine
learning due to its flexibility and inherent ability to describe uncertainty in
function estimation. In the context of control, it is seeing increasing use for
modeling of nonlinear dynamical systems from data,