Rafael’s research focuses on scientific machine learning that integrates physical knowledge into learning algorithms to improve their reliability and performance in engineering applications. He is particularly interested in neural operators for solving PDEs and enabling inverse design and optimization under physical constraints.
