Transfer learning from two-dimensional supercritical airfoils to three-dimensional transonic swept wings

Machine learning has been widely utilized in flow field modeling and aerodynamic optimization. However, most applications are limited to two-dimensional problems. The dimensionality and the cost per simulation of three-dimensional problems are so high that it is often too expensive to prepare sufficient samples. Therefore, transfer learning has become a promising approach to reuse well-trained two-dimensional models and greatly reduce the need for samples for three-dimensional problems. This pap...