The research project will be concerned with symmetries in classical mechanical systems, aiming to develop a machine learning algorithm using neural networks to identify them.

Noether’s theorem allows for the analytic identification of such symmetries but despite being proven over 100 years ago, no numerical implementation of the theorem has been developed that can find the symmetries in a system without having a prescribed list to choose from. This project aims to remove the requirement for a prescribed list to choose from, making the real-world implementation of such algorithms more feasible.