Fast Gaussian Process Methods for Point Process Intensity Estimation
Abstract
Point processes are difficult to analyze because they provide only a sparse and noisy observation of the intensity function driving the process. Gaussian Processes offer an attractive theoretical framework by which to infer optimal estimates of these underlying intensity functions. The result of this inference is a continuous function defined across time that is typically more amenable to analytical efforts. However, a naive implementation of this intensity estimation will become computationally infeasible in any problem of reasonable size, both in memory and run-time requirements. We demonstrate problem specific methods for a class of renewal processes that eliminate the memory burden and reduce the solve time by orders of magnitude.
Cite
Text
Cunningham et al. "Fast Gaussian Process Methods for Point Process Intensity Estimation." International Conference on Machine Learning, 2008. doi:10.1145/1390156.1390181Markdown
[Cunningham et al. "Fast Gaussian Process Methods for Point Process Intensity Estimation." International Conference on Machine Learning, 2008.](https://mlanthology.org/icml/2008/cunningham2008icml-fast/) doi:10.1145/1390156.1390181BibTeX
@inproceedings{cunningham2008icml-fast,
title = {{Fast Gaussian Process Methods for Point Process Intensity Estimation}},
author = {Cunningham, John P. and Shenoy, Krishna V. and Sahani, Maneesh},
booktitle = {International Conference on Machine Learning},
year = {2008},
pages = {192-199},
doi = {10.1145/1390156.1390181},
url = {https://mlanthology.org/icml/2008/cunningham2008icml-fast/}
}