A Smarter Way to Read the Stars: Developing the Next Generation of Stellar Classifiers

Description
Asteroseismology is the study of tiny vibrations inside stars. These vibrations reveal the hidden structure of a star in the same way that earthquakes reveal the interior of Earth. It is one of the most powerful tools in modern astrophysics, allowing us to measure stellar masses and radii with remarkable precision. The field is booming because telescopes such as Kepler and TESS have recorded thousands of these stellar vibration patterns.

Inspired by an existing neural network that evaluates the reliability of automated seismic measurements (Reyes et al., MNRAS 511, 5578–5596, 2022), this project aims to extend its current performance with red-giant branch stars to achieve a comparable level with red clump giants, helium-core burning stars with more complex power-spectrum patterns. The project will upgrade that classifier using modern machine learning methods, stronger algorithms and improved feature extraction. The model will be trained on data from the Kepler and TESS missions and designed to work across many types of stars. A key goal is to make the tool flexible, easy to retrain and capable of identifying red clump stars directly.

Image

What the student will do

The student will upgrade an existing seismic classifier using modern machine learning techniques and expanded training sets. They will develop a more robust, easily retrainable system that performs well even for challenging stars like those in the red clump. Working with Kepler and TESS data, they’ll build and test a classifier that can quickly process large stellar samples and deliver clean, reliable measurements for ongoing and next-generation Galactic archaeology studies.

Why this project is important

Reliable seismic measurements let astronomers “listen” to stars and figure out their sizes, ages, and evolutionary stages. When we can trust these measurements across huge datasets like those from Kepler and TESS, we unlock sharper views of how the Milky Way formed and evolved. Improving these tools means better maps of our Galaxy and stronger foundations for future research.