A unified scalable framework for causal sweeping strategies for Physics-Informed Neural Networks (PINNs) and their temporal decompositions

Michael Penwarden, Ameya D. Jagtap, Shandian Zhe, George Em Karniadakis, Robert M. Kirby

Abstract: Highlights•Unified framework to describe existing and new PINN causality methods.•Identification of distinct training challenges in temporal PINN decomposition.•Comparative computational speed-ups to existing and unmodified PINN methods.•Improved predictive capabilities on problems known to pose PINN training challenges.