Gaussian Processes for Regression
Abstract
The Bayesian analysis of neural networks is difficult because a sim(cid:173) ple prior over weights implies a complex prior distribution over functions . In this paper we investigate the use of Gaussian process priors over functions, which permit the predictive Bayesian anal(cid:173) ysis for fixed values of hyperparameters to be carried out exactly using matrix operations. Two methods, using optimization and av(cid:173) eraging (via Hybrid Monte Carlo) over hyperparameters have been tested on a number of challenging problems and have produced excellent results.
Cite
Text
Williams and Rasmussen. "Gaussian Processes for Regression." Neural Information Processing Systems, 1995.Markdown
[Williams and Rasmussen. "Gaussian Processes for Regression." Neural Information Processing Systems, 1995.](https://mlanthology.org/neurips/1995/williams1995neurips-gaussian/)BibTeX
@inproceedings{williams1995neurips-gaussian,
title = {{Gaussian Processes for Regression}},
author = {Williams, Christopher K. I. and Rasmussen, Carl Edward},
booktitle = {Neural Information Processing Systems},
year = {1995},
pages = {514-520},
url = {https://mlanthology.org/neurips/1995/williams1995neurips-gaussian/}
}