Ph.D. in Applied Mathematics (Statistics)
Ph.D. in Applied Mathematics (Statistics)

About me


Actually, I’m a research engineer  at the Labsticc laboratory (IMT-Atlantique Brest) in the MOTEL team.  

We are working on the amelioration of  LabNbook which is an online notebook for writing and learning sciences. More precisely, my job consists in the quantification and the evaluation of the collaborative work by creating dashboards and machine learning algorithms. Before that, I was a teacher and researcher at INSA de Rouen Normandie in the LITIS laboratory (team Doccapp). My main research interests was centered around:

  • Decision theory and estimation theory.
  • High dimensional Statistics.

You can find an overview of my research  here. The courses I have taught are:

  • Machine learning
  • Inferential Statistics
  • Descriptive Statistics
  • Matrix Computation

Bref Bio

I obtained my Ph.D in applied Mathematics at Université de Rouen Normandie (France) and at École Nationale Supérieure de Statistique et d’Économie Appliquée (ENSSEA, Algeria) in 2019. I was supervised by Dominique Fourdrinier and Fatiha Mezoued.  My thesis was Funded by the Franco-Algerian program PROFAS B+ 2016. During my thesis, I worked on high-dimensional Covariance matrix estimation in a linaire regression model. My dissertation is available here.



My main research interests center around estimation of the covariance matrix under decision theoretic point of view . In my published papers, I proposed estimators improving classical (unbiased) ones, used as reference estimators. The improvement is measured with respect to a loss function and its associated risk. I consider the large class of elliptically symmetric distribution. Such estimators can be used in the multivariate linear regression model.





INSA Rouen Normandy 2019-2021​

  • Introduction to machine learning with Gilles Gasso:
  • Statistics with Bruno Portier:
    • Times series,
    • Linear regression (simple and multiple)
    • ANOVA1 and ANOVA2
  • Inferential statistics with Bruno Portier:
    • Point estimation
    • Interval of confidence
    • Hypothesis testing
  • Matrix computation Gilles Gasso and Benoit Gaüzère
    • Matrix Factorizations : LU, LDL and Cholesky factorization, QR factorization
    • Solving  linear system
    • Review of Least squares problem
    • Application of matrix factorizations
    • Iterative methods (Jacobi, Gauss-Seidel, relaxation)
    • Algorithms for Eigenvalues and Singular values computation
  • Introduction to inferential statistics with Alexandrina Rogozan: Practical work with Python on inferential statistics (TP1, TP2, TP3, TP4, TP5, TP6 )
  • Updating in Maths with Benoit Gaüzère: Linear algebra, real analysis and complex numbers.

Université virtuel du Senegal Feb. 2021 (UVS):

Project in machine learning with Pyhton (24Hours): It consist on the implementation of the following approach : Multipe linaire regression, Ridge, variable selection, Lasso, Elastic-net. The objectif is to use theses methods (with python) in order to predict the price of an appartement. [TP] [data set]


Ads detection in TV on real time:

With the Python Library OpenCV, I am developing a program that detects the start and end time of a given ad. This project is independent of my academic research (I do it for Fun !). The code is available here

Covariance Matrix estimation:

Some codes for estimators of the covariance matrix for elliptical symmetric distributions are available here.

Portofolio selection with singular covariance matrix:

Here is an application, to portfolio selection, of my last research with Wei Lu.