PointClouds

Materials related to equivariant ML methods for point clouds, developed at the CZS Summer School 2023 in Heidelberg, Germany.

Authors and contributors

• David W. Hogg (NYU) (MPIA) (Flatiron)
• Kate Storey-Fisher (NYU)
• Soledad Villar (JHU)
• Sebastian T. Gomez (HITS)
• [add your name here]

Projects

• Build a kNN regression method for point clouds.
• Test the properties of the OTT-jax package.
  • Verify equivariances.
  • Test dependence on hyper-parameters.
• Learn things about the information content of dark-matter halos.
• Compare feature-based, optimal-transport-based, and graph-based methods for point clouds.
• Visualize point-cloud data.
• [add your projects here]