Analyzing Galaxy Morphology and Orientation Using Machine Learning

Abstract

Project Summary Understanding galaxy shapes and orientations is essential for studying how structure forms and evolves in the universe. Traditionally, galaxy classification; spiral, elliptical, irregular, has relied on manual inspection, but modern sky surveys such as the Sloan Digital Sky Survey (SDSS) now provide millions of images that demand automated analysis. Machine learning offers a powerful means of identifying patterns in these large datasets. In this project I will research model types and develop and train a machine learning model, to classify galaxies by morphology and estimate their orientations such as the ellipticity and axis angles. The goal is to uncover statistical relationships between galaxy shapes, orientations, and spatial distributions.

Methods Data will be obtained from publicly available surveys such as SDSS DR17 and the Galaxy Zoo project. After image preprocessing, the model will be trained using Python libraries such as TensorFlow and Astropy. Statistical analysis tools such as NumPy, SciPy, and Matplotlib will then be used to test for correlations between morphology, orientation, and galactic environment. Model performance will be validated against human-classified samples to ensure reliability.

Expected Outcomes A functional machine learning model for galaxy classification Quantitative measures of orientation and alignment Visualizations and statistical summaries of morphological correlations A research poster or paper suitable for presentation at Carroll’s SURF festival

Significance

By combining astrophysics and data science, this project demonstrates how computational tools can deepen our understanding of the cosmos. The methods developed here could support future large-scale survey analysis while providing valuable undergraduate research experience in both physics and machine learning.