On-line Gaussian process modelling method has been implemented for the study of the ozone pollution in the air. The selected method is suitable in particular because it adapts model according to streaming measurement data to the complex dynamics and not entirely known mechanism in the background of ozone generation. The paper is a result of the project preparation phase and is published ahead of the official project start.
COBISS.SI-ID: 26629159
The Gaussian-process (GP) model is an example of a probabilistic, non-parametric model with uncertainty predictions. It can be used for the modelling of complex, non-linear systems and also for the identification of dynamic systems. In this chapter we propose a method for the sequential selection of streaming data so that the size of the active set remains constrained. Furthermore, for better adjustment of the model to the system the hyperparameter values are optimised as well. The viability of the proposed method is tested on data obtained from two, nonlinear, dynamic systems. The chapter is a result of the project preparation phase and is published ahead of the official project start.
COBISS.SI-ID: 26062375