Can Gaussian process inference suffer from high curvature regions and divergences? For a few number of RV’s, we could use pairplots and other tools to understand divergences. Are there such tools available for GP’s?
Can Gaussian process inference suffer from high curvature regions and divergences? For a few number of RV’s, we could use pairplots and other tools to understand divergences. Are there such tools available for GP’s?