Back to publications

Weak Constraint Gaussian Processes for Optimal Sensor Placement

Dur, Tolga H. and Arcucci, Rossella and Mottet, Laetitia and Molina-Solana, Miguel and Pain, Christopher and Guo, Yike
Journal of Computational Science 42, 101110 (2020)

Abstract:

We present a Weak Constraint Gaussian Process (WCGP) model to integrate noisy inputs into the classical Gaussian Process (GP) predictive distribution. This model follows a Data Assimilation approach (i.e. by considering information provided by observed values of a noisy input in a time window). Due to the increased number of states processed from real applications and the time complexity of GP algorithms, the problem mandates a solution in a high performance computing environment. In this paper, parallelism is explored by defining the parallel WCGP model based on domain decomposition. Both a mathematical formulation of the model and a parallel algorithm are provided. We use the algorithm for an optimal sensor placement problem. Experimental results are provided for pollutant dispersion within a real urban environment.

Links:

DOI: 10.1016/j.jocs.2020.101110
PDF: https://authors.elsevier.com/c/1aq7B6gytGmcRF

Bibtex:

@article{Dur2019,
  author = {Dur, Tolga~H. and Arcucci, Rossella and Mottet, Laetitia and Molina-Solana, Miguel and Pain, Christopher and Guo, Yike},
  title = {Weak Constraint Gaussian Processes for Optimal Sensor Placement},
  journal = {Journal of Computational Science},
  year = {2020},
  volume = {42},
  articleno = {101110},
  doi = {10.1016/j.jocs.2020.101110},
  comment = {https://authors.elsevier.com/c/1aq7B6gytGmcRF},
  timestamp = {24}
}