AI-Driven Large Scale Structure Inference from Next-Generation astrophysical datasets in the Multi-messenger Era

This project uses artificial intelligence to better understand the large-scale structure of the Universe. Cosmologists study the Universe using different observational probes, including galaxy surveys and catalogs of gravitational-wave events. Gravitational waves are tiny vibrations of spacetime sourced by powerful events such as two stellar black holes colliding. These objects act as important tracers of the underlying cosmic web of dark matter—the invisible structure that shapes how galaxies, black holes, and other cosmic objects are distributed. To obtain a reliable picture of the Universe and understand the physics governing it, accurate astrophysical models of dark-matter evolution and the distribution of its tracers are required. In this project, machine learning will be used to create realistic simulated catalogs of cosmic tracers, including gravitational-wave events, and to model how observational limitations affect what can be detected. By comparing simulated and realistic observation-like data, the project will help identify how observed patterns connect to the Universe’s underlying structure and the physics behind it. The work combines expertise from the University of Waterloo and the University of Cambridge and develops transferable AI tools for data-intensive science.

Faculty Supervisor:

Ghazal Geshnizjani

Student:

Partner:

University of Cambridge

Discipline:

Mathematics

Sector:

Education

University:

University of Waterloo

Program: