Astronomy Project

My first school group project.

Start date: 2023-08-29

Written by: Abdilahi

MNSU Galaxy Identification Project with JWST

Project Description:

Utilizing imaging and spectroscopy data from the James Webb Space Telescope (JWST), this project focuses on developing machine learning tools to automatically identify galaxies that have recently collided or are in the process of colliding. The identification process is based on galaxy morphologies in the imaging data.

Objective:

1. Develop Python-based software leveraging ML tools for identification with confidence notations.
2. Apply ML classification on the dataset available.
3. Create classification models for correlations between merger states and star formation intensity.
4. Evaluate model performance using popular information criteria metrics.
5. Deliver software suitably commented for dissemination to astrophysics colleagues.

Example Galaxy: Binary bridge post-merger galaxy observed with the JWST. The morphology indicates a recent merger with broadband NIRCam imaging at λ=1.15μm (a) and 1.50μm (b) provided. The 2D grism spectrum (c) provides insights into the star formation characteristics.

Galaxy Image

Attempted Machine Learning Models

Below are the model names and accuracy based on the Par5 training dataset.