I’m a French postdoctoral researcher interested in both prediction and causation. My research stands at the crossroads of these two worlds.

Many causal estimators have been developed, but little guidance exists about when to use (or not) them. The heart of my PhD was to compare the performances of such estimators in various contexts through simulations and how mixing them with machine learning. However, causal inference is not only a statistical/estimation problem. Identifiability of causal effects is also a crucial challenge in modern research. Thereby, my current research aims to develop a tool to check positivity in various contexts (e.g., transportability, mediation, or longitudinal settings) non-parametrically. Machine learning is also a powerful tool for prediction. Alternatively, I investigate the potential of the super learner to improve the transportability of prediction models or how to personalize them for a given patient.

I’m supported by a CRM-StatLab-CANSSI postdoctoral fellowship.

Interests: Causal inference, Methodology, Prediction, R Programming, Simulation study, Super learning

Education:

   Ph.D. Biostatistics, Nantes Université, France

   M.Sc. Biostatistics, Nantes Université, France

   B.Sc. Physiology, Nantes Université, France