About Me
I am a recent graduate with a Master's in Data Science, actively seeking PhD opportunities in machine learning. I aspire to contribute to research developing adaptive, scalable systems where distributed agents can collaboratively learn and solve complex problems.
"The measure of intelligence is the ability to change." — Albert Einstein
Research Interests
Current Research Focus
My research focuses on the intersection of multi-agent systems, reinforcement learning, and lifelong learning, particularly in non-stationary environments. I'm interested in how distributed AI systems can collaboratively adapt, learn continuously, and transfer knowledge without centralized control.
My work has explored:
- One-shot learning techniques for object recognition in design thinking environments
- Gradient bandit algorithms for reinforcement learning in non-stationary reward distributions
- Exploration-exploitation trade-offs under shifting reward scenarios
I'm particularly interested in developing frameworks for adaptive collective intelligence that can enhance applications in disaster response, swarm robotics, and privacy-preserving collaborative systems.
Education and Academic Background
Master's in Data, Knowledge, and Intelligence
- Ranked first in cohort (Major de promotion)
- With Distinction (Très Bien)
- Circle-U Exchange Program at Universität Wien (2023-2024)
Bachelor of Computer Science
DUT in Multimedia and Internet Professions
Research Experience in Artificial Intelligence
Master's Thesis: One-Shot Object Recognition techniques for Digital Twins
- Created a flexible ResNet50Encoder-based framework for one-shot learning of new objects without retraining
- Supervised by Prof. D. Karagiannis (Universität Wien) and Assoc. Prof. M. Ouziri (Université Paris Cité)
Gradient Bandit Algorithms for Non-Stationary Environments
- Investigated exploration-exploitation trade-offs under shifting reward distributions
- Analyzed parameter optimization for reinforcement learning algorithms in dynamic environments
- Compared baseline and step-size effects on performance in both random walk and periodic reset scenarios
- Mathematics of Vision and Reinforcement Learning Research Project
Bachelor Thesis: Prey-Predator Multi-Agent Simulation
- Developed Unity-based ecosystem with emergent agent behaviors
- Implemented a logic-based agent AI
Academic Publications
One-shot learning for flexible object recognition in design thinking workshops.