Laboratoire CRIStAL (SigMA Team), Villeneuve-d'Ascq

Topic: Asynchronous MCMC algorithms for fast Bayesian inference

Developing new MCMC algorithms that are fast in high dimension and distributed (or even asynchronous).

Supervisors: Pierre-Antoine Thouvenin (Maître de conférence), Pierre Chainais (Professeur)

University of Lille, Villeneuve-d'Ascq

Double degree with École Centrale de Lille

• Advanced Mathematics: differential calculus & optimization, measure theory, linear algebra
• Machine Learning: kernel machines, deep learning (CNN, MLP, GANs, VAE), Bayesian approaches (MCMC), reinforcement learning, ethical aspects of AI, privacy
• Practice: OOP, parallelization, heuristic algorithms, research projects, data challenge

École Centrale de Lille, Villeneuve-d'Ascq

Course mainly oriented towards IT and management. Special attention to data science and programming (Android, JAVA).

• RASH project: Android programming of a GPS adapted for people with reduced mobility on the Lille campus
• IA and Health module: optimisation of an emergency patient file
• MOOC Project Management by R. BACHELET: national diploma (classical and advanced courses)

Carnot Highschool, Dijon

Preparatory classes for competitive exams for engineering schools specialized in mathematics, physics and chemistry.

Charles de Gaulle International Highschool, Dijon

A-level obtained in Science with highest honors and European mention.

NOCE Team, 6-month internship

• Study of Knowledge Tracing models based on the Self-Attention mechanism
• Articles publication for several related events like EIAH & IA
• Presentation at meetings like I2C in CRIStAL

4-month internship

Project: Elaboration of a model to predict money laundering

• Gathering and consolidation of external data
• Descriptive analysis and feature selection
• Elaboration of the first XGBoost models regarding several measures (AUC-PR, Accuracy, Gini...)
• Community Management about Data

4-week internship

• In-depth analysis of official and technical documents
• Study and collection of information from field staff
• Production of a summary document for maintenance agents