Sensor Placement for Spatial Gaussian Processes with Integral Observations

Krista Longi, Chang Rajani, Tom Sillanpää, Joni Mäkinen, Timo Rauhala, Ari Salmi, Edward Haeggström, Arto Klami
Proceedings of the 36th Conference on Uncertainty in Artificial Intelligence (UAI), PMLR 124:1009-1018, 2020.

Abstract

Gaussian processes (GP) are a natural tool for estimating unknown functions, typically based on a collection of point-wise observations. Interestingly, the GP formalism can be used also with observations that are integrals of the unknown function along some known trajectories, which makes GPs a promising technique for inverse problems in a wide range of physical sensing problems. However, in many real world applications collecting data is laborious and time consuming. We provide tools for optimizing sensor locations for GPs using integral observations, extending both model-based and geometric strategies for GP sensor placement.We demonstrate the techniques in ultrasonic detection of fouling in closed pipes.

Cite this Paper


BibTeX
@InProceedings{pmlr-v124-longi20a, title = {Sensor Placement for Spatial Gaussian Processes with Integral Observations}, author = {Longi, Krista and Rajani, Chang and Sillanp\"{a}\"{a}, Tom and M\"{a}kinen, Joni and Rauhala, Timo and Salmi, Ari and Haeggstr\"{o}m, Edward and Klami, Arto}, booktitle = {Proceedings of the 36th Conference on Uncertainty in Artificial Intelligence (UAI)}, pages = {1009--1018}, year = {2020}, editor = {Jonas Peters and David Sontag}, volume = {124}, series = {Proceedings of Machine Learning Research}, month = {03--06 Aug}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v124/longi20a/longi20a.pdf}, url = { http://proceedings.mlr.press/v124/longi20a.html }, abstract = {Gaussian processes (GP) are a natural tool for estimating unknown functions, typically based on a collection of point-wise observations. Interestingly, the GP formalism can be used also with observations that are integrals of the unknown function along some known trajectories, which makes GPs a promising technique for inverse problems in a wide range of physical sensing problems. However, in many real world applications collecting data is laborious and time consuming. We provide tools for optimizing sensor locations for GPs using integral observations, extending both model-based and geometric strategies for GP sensor placement.We demonstrate the techniques in ultrasonic detection of fouling in closed pipes. } }
Endnote
%0 Conference Paper %T Sensor Placement for Spatial Gaussian Processes with Integral Observations %A Krista Longi %A Chang Rajani %A Tom Sillanpää %A Joni Mäkinen %A Timo Rauhala %A Ari Salmi %A Edward Haeggström %A Arto Klami %B Proceedings of the 36th Conference on Uncertainty in Artificial Intelligence (UAI) %C Proceedings of Machine Learning Research %D 2020 %E Jonas Peters %E David Sontag %F pmlr-v124-longi20a %I PMLR %P 1009--1018 %U http://proceedings.mlr.press/v124/longi20a.html %V 124 %X Gaussian processes (GP) are a natural tool for estimating unknown functions, typically based on a collection of point-wise observations. Interestingly, the GP formalism can be used also with observations that are integrals of the unknown function along some known trajectories, which makes GPs a promising technique for inverse problems in a wide range of physical sensing problems. However, in many real world applications collecting data is laborious and time consuming. We provide tools for optimizing sensor locations for GPs using integral observations, extending both model-based and geometric strategies for GP sensor placement.We demonstrate the techniques in ultrasonic detection of fouling in closed pipes.
APA
Longi, K., Rajani, C., Sillanpää, T., Mäkinen, J., Rauhala, T., Salmi, A., Haeggström, E. & Klami, A.. (2020). Sensor Placement for Spatial Gaussian Processes with Integral Observations. Proceedings of the 36th Conference on Uncertainty in Artificial Intelligence (UAI), in Proceedings of Machine Learning Research 124:1009-1018 Available from http://proceedings.mlr.press/v124/longi20a.html .

Related Material