Skip to content
Ousmane Touat
PhD Student & Research Engineer
Specializing in Decentralized Learning, Federated Learning security, and Graph Neural Networks.
Experience
PhD Student
Sep. 2024 - PresentLIRIS/CNRS - INSA Lyon
Researching Byzantine resilience and robustness against external attacks in decentralized learning.
Python Software Engineer
Sep. 2023 - Sep. 2024Worldgrid France
Full-stack development for a major player in the French energy market using FastAPI, VueJS, Pandas, SQLAlchemy, and Openshift.
Research Engineer
Jun. 2022 - Jun. 2023LIRIS/CNRS - INSA Lyon
Researched security and fairness of Federated Learning algorithms and authored conference publications.
Research Intern
Jun. 2019 - Aug. 2019CITI Lab - INSA Lyon
Worked on very low latency audio processing on FPGA by high-level synthesis within the Syfala project.
Education
Master of Science in Computer Science
Sep. 2020 - Sep. 2022Universitat Passau
M.Eng in Computer Science
Sep. 2018 - Sep. 2022INSA Lyon
Undergraduate Study (CPGE MPSI/PSI)
Sep. 2016 - Jul. 2018Lycee Paul Eluard
Selected Publications & Projects
- Exposing the vulnerability of decentralized learning to membership inference attacks through the lens of graph mixing — Proceedings of the 26th International Middleware Conference (Middleware '25)
- Bias mitigation in federated learning for edge computing — Proc. ACM Interact. Mob. Wearable Ubiquitous Technol. (IMWUT)
- Gran is superior to graphrnn: Node orderings, kernel- and graph embeddings-based metrics for graph generators — Machine Learning, Optimization, and Data Science (LOD 2023)
- Towards robust and bias-free federated learning — Proceedings of the 3rd Workshop on Machine Learning and Systems (EuroMLSys '23)