Open Positions
Published:
I am regularly looking for motivated students and researchers to join us, working on trustworthy machine learning, world models, and their application to safety-critical systems, in particular medical imaging and neuro-imaging. Open positions are listed below by type.
Postdoctoral researcher
No open positions at the moment. Feel free to reach out at salah.ghamizi[at]uni.lu if you are interested in future opportunities.
PhD student
This PhD position focuses on designing novel neuro-symbolic learning encoders for world models applied to breast cancer (with CHL). The candidate will investigate architectures that embed mathematical models of breast tumor growth and treatment response into multimodal world models combining imaging, tabular clinical data, and graph-structured information, so that predictions remain consistent with known tumor biology and clinical constraints.
This PhD position focuses on designing novel neuro-symbolic learning encoders for world models applied to power grid systems. The candidate will investigate architectures that embed the electrical laws governing power flows (e.g. Kirchhoff’s laws, power balance constraints) and battery chemical dynamics directly into the learning process, so that the resulting world models remain physically consistent and generalize better than purely data-driven approaches.
Internship
Evaluate and improve generative models that synthesize PET from MRI for Alzheimer’s disease staging and biomarker quantification.
Build a multimodal foundation model for power grids that simulates scenarios, explains grid state, and advises operators on control actions.
Identify which medical imaging tasks can be combined, comparing multi-task learning and mixture-of-experts strategies.
Build a retrieval-augmented, privacy-preserving deployment pipeline for foundation models generating MRI scans across hospitals.
Jointly learn where to place grid sensors and how to estimate grid state robustly when sensors fail, using graph neural networks.