Abstract
Modeling physiological signals is challenging due to the complexity of systems governed by nonlinear dynamics and uncertain parameters.
This presentation explores the application of Physics-Informed Neural Networks (PINNs) to address these challenges, focusing on their dual capability for solving both forward and inverse problems.
The core strength of PINNs lies in their ability to embed the governing physical laws typically expressed as (PDEs) or (ODEs) directly into the neural network's loss function.
This approach ensures the resulting solutions are both consistent with observed data and faithful to the known physics. We will focus here on a problem coming from Neuroscience.