Equivariant Neural Networks for Classification of Dark Matter Substructure

The study of substructures in the dark matter has shown signs of promise to deliver on the open-ended problem of the identity of dark matter. Strong gravitational lensing has been proved to be a strong probe for dark matter substructure.

Current convolutional neural networks are only capable of translational equivariance. However, in a number of application (including ours), a larger groups of symmetries, including rotations and reflections are present in the data as well that needs to be exploited. This gives rise to the notion of Equivariant Convolutional Networks.

This project is the follow up work of the previous attempts to study and determine the morphology of dark matter substructure using deep learning based approaches.

Read more about the project on my Medium article.