About me and my background
About me and my background
If science is an adventure, then I am an adventurer. What is greater than understanding the universe? I am on this path, every day I move forward...
|Research, learning and teaching|
I am a PhD student working on a clustering algorithm, the goal is to improve the performance by using more sophisticated optimization methods.
[2020 - 2024 ] Fast methods for computing the Mahalanobis distance in automatic classification – Thesis
In data partitioning, the Mahalanobis distance improves the similarity analysis of individuals. Development of new automatic classification methods (FCM, ECM variants) using robust mathematical optimization methods (ADMM, Nesterov) allowing the use of this distance. Programming in Matlab.
MORE ABOUT MY THESIS ![2020] Stock price predictions – 3th year project
The prediction of stock prices is a delicate task because of the noise and uncertainties present in the information. Development of a method based on a neural network (LSTM) and a data preprocessing. Programming in Python with the Tensorlfow package.
[2019] Modeling of meat oxidation – 2nd year internship
Within INRA, analysis of the physico-chemical phenomena undergone by the meat and translation into a differential equation system. Solving the equation with an adapted Matlab solver. Simulation of different meat marketing routes and analysis of the results.
[2019] Study of the PointNet neural network – 3th year project
Understanding of the first neural network processing directly a 3D point cloud. Evaluation with Lidar data in the context of autonomous driving. First experience with deep learning and use of Python with the Tensorflow package.
[2020 - ] Doctoral School of Engineering Sciences (ED SPI) Clermont Auvergne University
Computer science.
[2017 - 2020] Institut Supérieur d'Informatique, de Modélisation et de leurs Applications (ISIMA) – Clermont-Ferrand
Mathematical modeling and data science.
[2015 - 2017] Prépa Maths Physique Option Informatique – Clermont-Ferrand
[2012 -2015] Filière Scientifique - Bac mention Très Bien Highschool Godefroy de Bouillon – Clermont-Ferrand
Pratical works :
Structure de données (Data structure) [Engineering school]
Learn to manage data in lists, stacks, queues... in C with pointers.
Numerical method application [Bachelor University]
Apply basic numerical method : solve a linear system, inverse a matrix, polynomial interpretation, linear regression.
Probabilties and Statistics [Bachelor University]
Studies of the principal laws of probabilities, continuous and discrete. Application of linear regression methods.
Lesson :
Calcul différentiel (Differential Calculus) [Engineering school]
Integration on curves and surfaces & Polynomial interpolation, numerical integration, numerical solution of differential equations.
End-of-study internship (master 2) :
A new performance Measure for Clustering with Mahalanobis distance
Development of a new measure to evaluate the quality of a score taking into account the Mahalanobis distance.
Engineering school project (licence 3) :
Démonstration d’active learning sur une variante de k-means
To improve the performance of a variant of k-means (CECM), we used constraint pairs allowing semi-supervised learning. The added constraints specify the area of uncertainty that has been previously delimited.
Handitutorat
Sopra Steria's project to help young people with disabilities obtain their baccalaureate by support from engineering school students. I had the chance to be accompanied in high school and then I had the opportunity in engineering school to participate in my turn.
Course de la diversité
For 10 years now, the Fédération Française du Sport d'Entreprise has been promoting diversity, cohesion and sharing through the Course de la Diversité, an event open to everyone and bringing together companies, local authorities and associations.
I took part in the 3km race in a time of 16 min 30 sec.