Artificial Intelligence Approaches to Dark Matter Particle Discovery
Name of applicant
Martin Rosenlyst Jørgensen
Title
Postdoctoral Fellow
Institution
The Massachusetts Institute of Technology (MIT)
Amount
DKK 1,437,275
Year
2024
Type of grant
Internationalisation Fellowships
What?
We aim to discover the particle properties and nature of Dark Matter through a novel approach combining particle physics modelling and supercomputer simulations of these models with Artificial Intelligence (AI) based data analysis. This would solve one of the most profound problems with our current Standard Model (SM) of particle physics.
Why?
The SM is our most fundamental theory of Nature but it is severely incomplete. It does not explain the origin of 85% of the mass in our Universe – the unknown Dark Matter (DM). The existence of DM is firmly established in astrophysical systems like galaxies but lab experiments have been unable to detect DM particles for half a decade. We need a complementary approach to particle discovery.
How?
Near the centers of galaxies, DM densities become so high that DM can collide, leaving observable imprints of their interactions but disturbed by astrophysical “noise”. We will develop realistic supercomputer simulations of these processes based on fundamental DM interactions and train AI-based algorithms on these simulations to detect these particle interactions in real noisy data.