Jakub Lengiewicz

Abstract: Physics-Informed Neural Networks (PINNs) are a powerful class of deep learning models that can effectively learn complex physical phenomena by integrating data with known physical equations and constraints. PINNs have gained significant popularity in recent years, primarily due to their ability to rapidly and accurately model the behavior of complex systems, even when working with incomplete or noisy data. By leveraging both data-driven and physics-based approaches, PINNs offer a unique advantage over traditional machine learning models and have shown great potential in various fields, including physics, engineering, and finance. In this seminar, I will introduce the concept of PINNs and provide a theoretical basis for understanding how they work. I will discuss key challenges in developing PINNs, and explore current research trends in PINNs, including multi-fidelity modeling and transfer learning. This seminar is tailored for researchers who are interested in using PINNs but have no prior experience with them.

Dr. Mohammad Mahdi Rajabi is a post-doctoral researcher within the Faculty of Science, Technology and Medicine at the University of Luxembourg.