Gaussian Processes

ConjugateModel	A joint model with Gaussian likelihood: conditioning is exact.
HeteroscedasticModel	A joint model with input-dependent (heteroscedastic) noise.
JointModel	The joint distribution $p(f, y) = p(y mid f),p(f)$.
NonConjugateModel	A joint model with non-Gaussian likelihood.
Prior	A Gaussian process prior object.
construct_model	Construct the joint model for a prior/likelihood pair.

Conditioning

ExactPosterior	Exactly conditioned GP: a Gaussian likelihood integrated analytically.
LatentPosterior	Approximately conditioned GP for non-Gaussian likelihoods.
Posterior	A conditioned Gaussian process, \(p(f \mid \mathcal{D})\).