Gaussian Process Regression

[sourish-cmi]

Integrate GP Regression from GaussianProcesses.jl with CRRao.jl

Integrate GP Regression from GaussianProcesses.jl with CRRao.jl

Ref
[1] Fairbrother, Jamie etal, "GaussianProcesses. jl: A Nonparametric Bayes Package for the Julia Language", (2022) Journal of Statistical Software, 102, 1--36,

The possible solution would look like

container = fit(Formula, data::DataFrame, modelClass::GaussianProcessRegression,IndexVar,MeanFunction::MeanZero,kernelClass::Exponential, DistanceClass::Eucledian)

Example

container = fit(y~x1+x2+x3,train_data,z,GaussianProcessRegression(),MeanZero(),Exponential())

Here we are trying to fit
y ~ GP(mu(x),Sigma(z)), where z is the index variable or distance would be estimated using z.

Spatial Statistics

container = fit(y~x1+x2+x3,train_data,s,GaussianProcessRegression(),MeanZero(),Exponential(),EarthDistance())

In the above code s contain lat and lon.