Area Machine Learning and Simulation

Topic Applications of Machine Learning and Simulation in Gravitational Lensing

This topic is an on-going, inter-disciplinary endeavour, aiming to map the dark matter of the universe. Emitting no light, dark matter can only be observed indirectly, as gravity deflects light from more distant galaxies.

The project would aim to use machine learning to reconstruct the distribution of dark matter from distorted images of remote galaxies.

There are several options for the specialisation module: (1) Machine Vision adapting the level 3 module on the topic, (2) a syllabus in physics/cosmology, or even (3) a syllabus in image processing. The latter is particular useful for working with empirical data.

Team supervision. Project students join the team, including students at different levels and degree programmes, as well as academic staff. Individual supervision is provided as required.

Supervisors: Hans Georg Schaathun (Computing/Machine Learning) and Ben David Normann (Physics/Cosmology)

Assessment: project report/portfolio for the project and viva voce for the specialisation module.

We seek one student from S&V, in addition to one student doing her thesis in physics. With exceptional motivation, it may be possible to accommodate two students from S&V.

This topic is well-suited for continuation into a full dissertation and scientific publication.