A Bayesian neural network approach to Multi-fidelity surrogate modelling

Abstract

This paper deals with surrogate modelling of a computer code output in a hierarchical multi-fidelity context, i.e., when the output can be evaluated at different levels of accuracy and computational cost. Using observations of the output at low- and high-fidelity levels, we propose a method that combines Gaussian process (GP) regression and Bayesian neural network (BNN), in a method called GPBNN. The low-fidelity output is treated as a single-fidelity code using classical GP regression. The high-fidelity output is approximated by a BNN that incorporates, in addition to the high-fidelity observations, well-chosen realisations of the low-fidelity output emulator. The predictive uncertainty of the final surrogate model is then quantified by a complete characterisation of the uncertainties of the different models and their interaction. GPBNN is compared with most of the multi-fidelity regression methods allowing to quantify the prediction uncertainty.

References in corpus (8)

• Hands-on Bayesian Neural Networks -- a Tutorial for Deep Learning Users
• A composite neural network that learns from multi-fidelity data: Application to function approximation and inverse PDE problems
• Pyro: Deep Universal Probabilistic Programming
• Deep UQ: Learning deep neural network surrogate models for high dimensional uncertainty quantification
• Multi-fidelity Bayesian Neural Networks: Algorithms and Applications
• Nested Kriging predictions for datasets with large number of observations
• Multi-Fidelity Bayesian Optimization via Deep Neural Networks
• A graphical multi-fidelity Gaussian process model, with application to emulation of heavy-ion collisions