Sub-modules for learning without derivatives.
module Spec = Spec.EvalSpecification of covariance function
module Inducing = FITC.Eval.InducingEvaluating inducing inputs
module Input = FITC.Eval.InputEvaluating single inputs
module Inputs = FITC.Eval.InputsEvaluating (multiple) inputs
module Model = FITC.Eval.Model(Untrained) model - does not require targets
module Trained = FITC.Eval.TrainedTrained model - requires targets
module Stats : sig ... endStatistics derived from trained models
module Mean_predictor = FITC.Eval.Mean_predictorModule for making mean predictions
module Mean = FITC.Eval.MeanPosterior mean for a single input
module Means = FITC.Eval.MeansPosterior means for (multiple) inputs
module Co_variance_predictor = FITC.Eval.Co_variance_predictorModule for making (co-)variance predictions
module Variance = FITC.Eval.VariancePosterior variance for a single input
module Variances = FITC.Eval.VariancesPosterior variances for (multiple) inputs
module Covariances : sig ... endPosterior covariances
module Sampler = FITC.Eval.SamplerModule for sampling single points from the posterior distribution
module Cov_sampler : sig ... endModule for sampling (multiple) points from the posterior distribution accounting for their covariance