Latent variable Gaussian process models with fully Bayesian inference.

LVGP-Bayes is a Python library for estimating Latent variable Gaussian process (LVGP) models through fully Bayesian inference. This respository contains code to run the experiments in the paper Fully Bayesian inference for latent variable Gaussian process models.

For reproducing the experiments, refer to the each subdirectory in the experiments/ folder.

Note: The code is under an Academic and Non-Commerical Research use license.

Installation

git clone https://github.com/syerramilli/lvgp-pytorch <path>
pip install <path>

Note: <path> is optional.

Citing

@article{yerramilli2023fully,
  title={Fully Bayesian Inference for Latent Variable Gaussian Process Models},
  author={Yerramilli, Suraj and Iyer, Akshay and Chen, Wei and Apley, Daniel W},
  journal={SIAM/ASA Journal on Uncertainty Quantification},
  volume={11},
  number={4},
  pages={1357--1381},
  year={2023},
  publisher={SIAM}
}