project 3

This project aims at using unsupervised learning techniques to find groupings of the Cannon (a mass stellar classifier for the APOGEE survey) dataset in chemical abundance space and determine if the same groupings also occur in physical space.

This project achieved the following:

• Found groupings of the Cannon dataset in chemical abundance space using Python Scikit-Learn’s K-means function.
• Conducted dimension reduction using Python Scikit-Learn’s UMAP function and found groupings of the Cannon dataset in chemical abundance space using GMM function.
• Found groupings of the Cannon dataset in chemical abundance space using Python Scikit-Learn’s DBSCAN function.

The project report is available here.