Gaussian Process-Based Real-Time Learning for Safety Critical Applications
Armin Lederer, Alejandro J Ordóñez Conejo, Korbinian A Maier, Wenxin Xiao, Jonas Umlauft, Sandra Hirche
Abstract
The safe operation of physical systems typically relies on high-quality models. Since a continuous stream of data is generated during run-time, such models are often obtained through the application of Gaussian process regression because it provides guarantees on the prediction error. Due to its high computational complexity, Gaussian process regression must be used offline on batches of data, which prevents applications, where a fast adaptation through online learning is necessary to ensure safety. In order to overcome this issue, we propose the LoG-GP. It achieves a logarithmic update and prediction complexity in the number of training points through the aggregation of locally active Gaussian process models. Under weak assumptions on the aggregation scheme, it inherits safety guarantees from exact Gaussian process regression. These theoretical advantages are exemplarily exploited in the design of a safe and data-efficient, online-learning control policy. The efficiency and performance of the proposed real-time learning approach is demonstrated in a comparison to state-of-the-art methods.
BibTeX
@InProceedings{pmlr-v139-lederer21a,
title = {Gaussian Process-Based Real-Time Learning for Safety Critical Applications},
author = {Lederer, Armin and Conejo, Alejandro J Ord{\'o}{\~n}ez and Maier, Korbinian A and Xiao, Wenxin and Umlauft, Jonas and Hirche, Sandra},
booktitle = {Proceedings of the 38th International Conference on Machine Learning},
pages = {6055--6064},
year = {2021},
editor = {Meila, Marina and Zhang, Tong},
volume = {139},
series = {Proceedings of Machine Learning Research},
month = {18--24 Jul},
publisher = {PMLR},
pdf = {http://proceedings.mlr.press/v139/lederer21a/lederer21a.pdf},
url = {https://proceedings.mlr.press/v139/lederer21a.html},
abstract = {The safe operation of physical systems typically relies on high-quality models. Since a continuous stream of data is generated during run-time, such models are often obtained through the application of Gaussian process regression because it provides guarantees on the prediction error. Due to its high computational complexity, Gaussian process regression must be used offline on batches of data, which prevents applications, where a fast adaptation through online learning is necessary to ensure safety. In order to overcome this issue, we propose the LoG-GP. It achieves a logarithmic update and prediction complexity in the number of training points through the aggregation of locally active Gaussian process models. Under weak assumptions on the aggregation scheme, it inherits safety guarantees from exact Gaussian process regression. These theoretical advantages are exemplarily exploited in the design of a safe and data-efficient, online-learning control policy. The efficiency and performance of the proposed real-time learning approach is demonstrated in a comparison to state-of-the-art methods.}
}