Multifidelity Bayesian Optimization: A Review

What is it about?

Multi-fidelity and Bayesian approaches are increasingly employed to tackle complex and costly engineering design optimization problems. This paper reviews recent advancements in multi-fidelity Bayesian optimization (MF BO), offering a clear and structured overview of its key components: Gaussian Process (GP) based multi-fidelity surrogate models and acquisition functions. We explore how MF BO can address various challenging topics in engineering design and identify promising areas for future research.

Featured Image