Physics-informed neural networks (PINNs) are a powerful tool that can solve partial differential equations (PDEs) and effectively model complex physical systems from data while also being informed by physical laws. However, the solution found by a PINN can violate conservation laws, which can prove to be unsafe if they are applied in a real-world setting. My research will investigate whether adding a projection layer that enforces a conservation invariant to the PINN will improve accuracy and conservation of some physical quantity.

I am participating in SuperUROP to learn how proper self-guided research is done. Over the past two summers, I have done ML and NLP research that have led me to seek a more formal and engaging research opportunity. I hope to continue to learn the skills that only come with doing real research in order to be more prepared if I decide to continue research and attend a graduate program.