Multi-fidelity surrogate model approach to optimization

Recently the use of Radial Basis Functions (RBF) has been introduced as an optional alternative to co-Kriging in the context of multi-fidelity surrogate modeling. In this paper, we compare the performance of Random Forest-based co-surrogates to the previously introduced co-Kriging and co-RBF using a set of bi-fidelity benchmark problems in 2, 4 and 8 dimensions. Our results show that there is a minimal overall difference between the different co-surrogate models with regards to final performace, although the training of Random Forests takes much less time compared to the Kriging and RBF methods.