PhD student · CMAP, École Polytechnique
Ayoub Oulad Ali
I work at the interface of applied mathematics and machine learning. My research focuses on physics-informed generative models and self-supervised learning for polarimetric Mueller imaging, with applications to cervical tissue analysis.
PhD in the IMOCEP project (PEPR Maths-Vives) on preterm birth risk and cervical cancer.
Research
Full research page →Physics-informed generative models
Deep generative architectures for Mueller matrices that respect positivity and realizability constraints, and produce useful synthetic data when clinical datasets are small.
Representation learning and uncertainty
Self-supervised pre-training (JEPA-style architectures), uncertainty-aware learning and interpretable latent representations for biomedical imaging.
Internship report (PDF, 116 pages): Mueller imaging with physics-constrained generative models.
Teaching, 2026–27
All courses →- Introduction to Machine Learning (ML1)Master X-HEC · tutorial sessions (PC), with course material
- Monte Carlo MethodsEngineering Cycle, Year 3 · practical sessions (TP), with corrections
- StatisticsEngineering Cycle, Year 2 · course page
Background
Full profile →PhD student at CMAP, École Polytechnique (since October 2026), after a research internship at CMAP (April–September 2026). MSc in Applied Mathematics at Sorbonne Université; BSc in Mathematics and Computer Science at Université de Nîmes. Winner of the ENS Data Challenge 2025.
Contact
Open to discussions on research collaborations in generative models, scientific machine learning and medical imaging.