Active Learning of Linear Embeddings for Gaussian Processes
Abstract
We propose an active learning method for discovering low-dimensional structure in high-dimensional Gaussian process (GP) tasks. Such problems are increasingly frequent and important, but have hitherto presented severe practical difficulties. We further introduce a novel technique for approximately marginalizing GP hyperparameters, yielding marginal predictions robust to hyperparameter mis-specification. Our method offers an efficient means of performing GP regression, quadrature, or Bayesian optimization in high-dimensional spaces.
Cite
Text
Garnett et al. "Active Learning of Linear Embeddings for Gaussian Processes." Conference on Uncertainty in Artificial Intelligence, 2014.Markdown
[Garnett et al. "Active Learning of Linear Embeddings for Gaussian Processes." Conference on Uncertainty in Artificial Intelligence, 2014.](https://mlanthology.org/uai/2014/garnett2014uai-active/)BibTeX
@inproceedings{garnett2014uai-active,
title = {{Active Learning of Linear Embeddings for Gaussian Processes}},
author = {Garnett, Roman and Osborne, Michael A. and Hennig, Philipp},
booktitle = {Conference on Uncertainty in Artificial Intelligence},
year = {2014},
pages = {230-239},
url = {https://mlanthology.org/uai/2014/garnett2014uai-active/}
}