Projects
Beyond my research activities, I have also worked on a number of non-academic projects, including a lot of group projects, some of which were carried out in collaboration with companies.
Forecasting CO2 Emissions in Real Estate - In collaboration with BNP Paribas, France
I chose to participate in this project alongside my Machine Learning course during my first year of master’s studies, as it provided a concrete and applied setting in which to implement the concepts we learned in class. The work lasted three months, with weekly meetings with data scientists from BNP Paribas. The project was conducted in a 4-person team, the focus being on the study of times series in order to forecast the future CO2 emissions by real estate property category. Tasks include data preprocessing, feature engineering, training models such as ARIMA, LSTM and PROPHET and performance benchmarking. You can find the git repository of the project below, and the report as weell.
Project repository
Project report (FR) (PDF)
Optimization of an electromagnetic anechoic chamber - MICS laboratory, CentraleSupélec, france
With a given wave propagation PDE model in a chamber, the goal was to minimize the energy of an electromagnetic wave in the chamber through determining the optimal distribution of absorbing material. The work involved both the theoretical study of the model (well-posedness of the PDE system) and the implementation of numerical algorithms; finite elements for solving the PDEs, and gradient descent in Hilbert spaces to determine the optimal distribution of absorbing lmaterial. You can find the report and repository of the project below.
Project repository
Project report (FR) (PDF)
Viral Marketing Campaign - In collaboration with Artefact, France
As part of a one month project with the data and AI consultancy Artefact, I integrated a four-person team analyzed the viral dynamics of an online climate-awareness campaign created for Greenpeace. The campaign involved a fictional real-estate startup (“Orizon”) promoting seaside homes projected for the year 2100, using IPCC and NASA climate data. Our work consisted in studying how information spread across social networks, identifying key influencers, modeling diffusion patterns, and proposing data-driven strategies to increase the campaign’s reach and impact. The project combined elements of graph analysis, social-network modeling, and practical decision-making for digital marketing. Below is the project repository.
Project repository
Coding Weeks - CentraleSupélec, France
As part of my education at CentraleSupélec, I participated in a two-week coding project in a five-person team. We got to choose the project we wanted to present, so me and my teammates decided to implement various user-customizable versions of Conway’s game of life. Then we finally implemented a small video game coded in Python, the principle of which relied on the dynamic of the game of life. Below is the project repository. You can play the game, by using the main.py file in the survival-the-game-of-life folder.
Project repository
Detection of anomalies in an IOT network - In collaboration with Bouygues-Télécom, France
In a 5-person team, I conducted an analysis of a log dataset following a DDoS attack to identify the characteristics of the attack. This project was realized as a conclusion to the Information theory/IOT course I followed during my first year of studies at CentraleSupélec. The tasks included developing machine learning tools for the detection of streaming DDoS attacks.