Project for designing and demonstrating physics-informed Gaussian process regression using data from an experimental optical network link.

Includes notebooks to generate the results from: J.W. Nevin, F.-J. Faquero Callabero, D. Ives and S.J. Savory, Physics-Informed Gaussian Process Regression for Optical Fibre Communication Systems, Journal of Lightwave Technology, 2021. https://doi.org/10.1109/JLT.2021.3106714

See my fork of SciKit Learn for my implementation of the physics-informed GP, specifically the additional method fit_phys in the GaussianProcessRegressor class.