Recently graduated in Applied Mathematics and Computer Science at Polytech Sorbonne, I am now specialising in Statistics at Sorbonne University. I am also looking for a research internship in the field of causal inference or inverse problem solving that could lead me to an academic thesis. See more about me.
Academic Interests
Causal Inference
Reading The Book of Why: The New Science of Cause and Effect (J.Pearl, 2018) recently introduced me to an absolutely fascinating field of research. At the junction of statistics and theoretical computer science, it interests me both in the questions it raises in our relationship to science and in the various applications it allows. So I am currently supplementing my knowledge on this subject by reading the books Causality (J.Pearl, 2009) and Element of Causal Inference (Pearl & al., 2017). As part of my PhD, I am actually particularly developping Causal Representation Learning methods for spatio-temporal data.
Earth System Sciences
I was introduced to climate science and the study of the Earth system as a whole in a course given by Hervé Le Treut during my preparatory classes. Since then, I have been interested in scientific issues related to earth system modelling, which is why I did my final year internship at the Locéan laboratory on the issues of tuning coupled ocean-atmosphere models using machine learning and Bayesian optimisation methods (History Matching). You may read my report for more information about the work I did there.I am currently pursuing this path by doing my PhD at the IPL where I am developing Causal Representation Learning methods for the Earth Sciences.
Kernel Methods
In the course of my career I have developed a great interest in Kernel methods, whether for their mathematical elegance, the diversity of methods they allow to develop or the variety of applications in which they are a precious tool. I am now working on the development of Causal Representation Learning methods relying mainly on notions of independence in Reproducing Kernel Hilbert Spaces (as it as been proposed in Fukumizu & al., 2004 or Cheng & al., 2022).
Nonparametric inference
Having discovered non-parametric inference during a statistical learning project, I was able to develop my knowledge of this field of research in statistics and data science during an introductory course on non-parametric methods supervised by Charlotte Dion-Blanc and Ismael Castillo), for which I carried out a project on the estimation of density and survival function by kernel methods within the framework of the multiplicative censoring model.
Academic projects
Read More →
Main Confounder Analysis for Teleconnection Patterns Analysis
Causal Representation Learning
|
Teleconnection Patterns Analysis
|
Cause-Effect Estimation
Read More →
History Matching for climate model tuning : experiments on the Lorenz-96
Climate Science
|
Climate Model Calibration
|
Inverse Problem Solving
|
Bayesian Optimisation
Read More →
Robust Machine Learning - study of the Median of Means estimator (Lecue & Lerasle, 2017)
Robust Estimation
|
Machine Learning
Read More →
Nonparametric density and survival function estimation - multiplicative censoring model (Comte & al, 2016)
Non-parametric Estimation
Read More →
Statistical learning for the detection of deficient MMR crypts to aid in the diagnosis of Lynch disease
Computer Vision
|
Deep Learning
|
One-Shot Learning
|
Machine Learning for Medical Imaging
Read More →
Online Convex Optimisation - Study of the Dual Averaging Methods (Xiao, 2009)
Machine Learning
|
Online Optimimisation
Read More →
High-Dimensional Regression - study of the debiased LASSO estimator (Javanmard & Montanari, 2013)
High-Dimensional Data
|
LASSO Estimator
|
Multivariate Hypothesis Testing
C.V.
Having completed my engineering degree in applied mathematics at Polytech Sorbonne, I am now specializing in Statistics at Sorbonne University in order to move towards an academic thesis…