bhetGP: Bayesian Heteroskedastic Gaussian Processes
Parul V. Patil · 2025
Performs Bayesian posterior inference for heteroskedastic Gaussian processes. Models are trained through MCMC including elliptical slice sampling (ESS) of latent noise processes and Metropolis-Hastings sampling of kernel hyperparameters. Replicates are handled efficientyly through a Woodbury formulation of the joint likelihood for the mean and noise process (Binois, M., Gramacy, R., Ludkovski, M. (2018) ) For large data, Vecchia-approximation for faster computation is leveraged (Sauer, A., Cooper, A., and Gramacy, R., (2023), ). Incorporates 'OpenMP' and SNOW parallelization and utilizes 'C'/'C++' under the hood.