Nicolas de Staël (1914 – 1955), Paysage du Lavandou, 1952
I am a professor at CentraleSupélec and a researcher at the Laboratoire des Signaux et Systèmes (L2S, UMR 8506). My work lies at the intersection of uncertainty quantification, statistical modelling, numerical simulation, machine learning and decision theory.
Research
My research focuses on Gaussian processes and sequential decision methods for problems in which experiments or numerical simulations are expensive. It addresses function approximation using Gaussian processes, Bayesian optimization, reliability analysis, set inversion and the calibration of predictive distributions. This work combines theoretical analysis, methodological developments, scientific software and applications in engineering.
I contribute to open-source scientific software. I currently develop GPmp, a Python package for Gaussian-process modelling, and gpmp-contrib, which provides methods for Bayesian optimization, excursion-set estimation and set inversion. I previously contributed to STK, a MATLAB/Octave toolbox for kriging.
Teaching
At CentraleSupélec, I coordinate the Pôle Projets Data Science, a project-based programme for first- and second-year engineering students. My teaching activities cover statistics, data science and artificial intelligence, including a course on signal and image compression and denoising. Since 2026, I have been responsible for the HubIA computing infrastructure.