Learning Physical Constraints with Neural Projections Shuqi Yang

We propose a new family of neural networks to predict the behaviors of physical systems by learning their underpinning constraints.