Skip to content

libKriging/saferGPMLE

 
 

Repository files navigation

saferGPMLE

License Last commit Lint

What you will find in here

This repository provides code and data to reproduce the experiments presented in:

Numerical issues in maximum likelihood parameter estimation for Gaussian process interpolation (2021)
Subhasish Basak, Sébastien Petit, Julien Bect, Emmanuel Vazquez
https://arxiv.org/abs/2101.09747

Table of contents

Citation

Please refer to this work using the following bibtex entry:

@misc{basak:2021:saferGPMLE,
  title        = {Numerical issues in maximum likelihood parameter
                  estimation for {G}aussian process interpolation},
  author       = {Basak, Subhasish and Petit, Sébastien and
                  Bect, Julien and Vazquez, Emmanuel},
  year         = {2021},
  howpublished = {arXiv:2101.09747},
  url          = {https://arxiv.org/abs/2101.09747}
}

Directory layout

There are two subdirectories in this repository:

  • pkgcomp provides a framework for comparing several Python toolboxes, as exemplified by Table 1 in the article.

  • safergpy provides a framework for reproducing the other results of the article, and in particular the results presented in Section 5.

Related resources

A presentation of this work is available on Zenodo: DOI:10.5281/zenodo.4653845.

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 96.8%
  • Jupyter Notebook 3.2%