SDSS MOC4 Asteroids' Color Classifications
A project designed to apply J. M. Carvano's correction methods to the Sloan Digital Sky Survey MOC4 asteroid database, and then use unsupervised learning to try to replicate his classifications. You can read the proposal here, or you could just jump to the conclusion by reading the final paper here.
The following unsupervised learning algorithms were used:
- K-Means++
- K-Medioids
- HDBSCAN*
- Agglomerative Clustering