Comparison of data reduction criteria for online Gaussian processes

Wietzke T, Graichen K (2026)


Publication Language: English

Publication Type: Journal article

Publication year: 2026

Journal

Book Volume: 62

Article Number: 101084

DOI: 10.1016/j.arcontrol.2026.101084

Abstract

Gaussian Processes (GPs) are widely used for regression and system identification due to their flexibility and ability to quantify uncertainty. However, their computational complexity limits their applicability to small datasets. Moreover, in a streaming scenario more and more datapoints accumulate which is intractable even for Sparse GPs. Online GPs aim to alleviate this problem by e.g. defining a maximum budget of datapoints and removing redundant datapoints. This work provides a unified comparison of several reduction criteria, analyzing both their computational complexity and reduction behavior. The criteria are evaluated on benchmark functions and real-world datasets, including dynamic system identification tasks. Additionally, extended insertion criteria are proposed to further filter out redundant datapoints. This work presents practical guidelines for choosing a suitable criterion for an online GP algorithm.

Authors with CRIS profile

Related research project(s)

How to cite

APA:

Wietzke, T., & Graichen, K. (2026). Comparison of data reduction criteria for online Gaussian processes. Annual Reviews in Control, 62. https://doi.org/10.1016/j.arcontrol.2026.101084

MLA:

Wietzke, Thore, and Knut Graichen. "Comparison of data reduction criteria for online Gaussian processes." Annual Reviews in Control 62 (2026).

BibTeX: Download