Abstract: In aerodynamic design, accurate and robust surrogate models are important to accelerate computationally expensive computational fluid dynamics (CFD)-based optimization. In this paper, a machine learning framework is presented ...