Supervisors: Michał Bejger, Valéria Carvalho, email: vcarvalho@camk.edu.pl
Title: What Can Gravitational Waves Waveforms Tell Us About Neutron Stars Through Machine Learning?
Timeline: 4 weeks (starting July) flexible
Description: The detection of gravitational waves (GWs), particularly from neutron star mergers such as GW170817, has opened a new window into the study of dense matter. The waveform of these signals encodes valuable information about the internal structure of neutron stars, including their equation of state, radius, and tidal deformability.
However, extracting this information using traditional methods can be computationally expensive and limited in flexibility. Modern machine learning (ML) techniques offer a promising alternative by enabling fast, data-driven inference from complex signals.
In this project, the student will explore how ML models can be used to extract physical properties of neutron stars from GW waveforms. The focus will be on understanding the connection between waveform features and neutron star physics, and on implementing simple ML pipelines to perform parameter estimation or classification tasks.
Work Plan:
Week 1 – Introduction & Setup: Learn the basics of neutron stars, gravitational waves, and machine learning. Get familiar with the dataset and Python tools.
Week 2 – Data Exploration: Explore and visualise the waveform data. Build simple models to understand the relationship between waveforms and neutron star properties.
Week 3 – Model Development: Improve the machine learning methods and test their performance.
Week 4 – Results: Analyse the results and discuss what was learned
Requirements: undergrads with interest and at least some knowledge of coding in Python, machine learning and statistics.