Scaling Multidimensional Gaussian Processes Using Projected Additive Approximations
Abstract
Exact Gaussian Process (GP) regression has O(N^3) runtime for data size N, making it intractable for large N. Advances in GP scaling have not been extended to the multidimensional input setting, despite the preponderance of multidimensional applications. This paper introduces and tests a novel method of projected additive approximation to multidimensional GPs. We thoroughly illustrate the power of this method on several datasets, achieving close performance to the naive Full GP at orders of magnitude less cost.
Cite
Text
Gilboa et al. "Scaling Multidimensional Gaussian Processes Using Projected Additive Approximations." International Conference on Machine Learning, 2013.Markdown
[Gilboa et al. "Scaling Multidimensional Gaussian Processes Using Projected Additive Approximations." International Conference on Machine Learning, 2013.](https://mlanthology.org/icml/2013/gilboa2013icml-scaling/)BibTeX
@inproceedings{gilboa2013icml-scaling,
title = {{Scaling Multidimensional Gaussian Processes Using Projected Additive Approximations}},
author = {Gilboa, Elad and Saatçi, Yunus and Cunningham, John and Gilboa, Elad},
booktitle = {International Conference on Machine Learning},
year = {2013},
pages = {454-461},
volume = {28},
url = {https://mlanthology.org/icml/2013/gilboa2013icml-scaling/}
}