Reduced Order Models for Non-Newtonian Fluids Using Non-Stationary Gaussian Process Regression
Published in , 2025
This (non-published) paper was written for team internship and explores how adaptive learning can be used to reduce the need for data in reduced order moddeling approaches for FEM simulations. The result is a python package for non-stationairy Gaussian Process Regression (NSGPR) and a method which needs a third of the data to gain the same accuracy.