Breaking the Baryonic Degeneracy II: Advanced Cosmological Inference and Cross-Simulation Robustness with CAMELS

Project Description:

While the foundational Honours project focuses on the initial detection of cosmological signals within a single simulation suite, this Master's level extension aims to address the critical challenge of model generalizability and precision inference. In the era of high-precision surveys, distinguishing the fundamental parameters of the Universe from the non-linear ``noise" of galaxy formation requires techniques that are both robust across different physical models and capable of providing rigorous uncertainty quantification. Building upon the previous work with the CAMELS dataset, this project shifts focus from simple parameter regression to a more comprehensive investigation of the ``Information Content" available in modern cosmological simulations. We seek to understand if the features learned by artificial intelligence represent universal physical laws or are contingent on specific numerical implementations of feedback.

Research Area:

Astrophysics

Project Level:

Masters

This Project Is Offered At The Following Node(s):

(NWU)

Special Requirements:

Candidates should have a strong proficiency in Python and an interest in computational astrophysics. Completion of the introductory project or equivalent experience in basic machine learning and simulation analysis is highly recommended.