About

About

Abdenour Soubih

Machine Learning Security Researcher | Federated Learning | Privacy-Preserving AI
Master’s Student in Computer Science & Engineering
Sungkyunkwan University (SKKU), South Korea 🇰🇷

👋 About Me

I am a Master’s researcher in Machine Learning Security, with a strong focus on federated learning, privacy leakage, and robustness under real-world constraints.
My work sits at the intersection of theory, systems, and experimentation, aiming to bridge practical deployment challenges with rigorous security analysis.

I have hands-on experience designing and evaluating federated learning systems under heterogeneous data, studying poisoning and inference attacks, and exploring privacy-preserving mechanisms such as Differential Privacy and secure auditing protocols.

Beyond research, I bring a solid engineering background in IoT systems, distributed platforms, and cloud-based infrastructures, gained through international academic and industry collaborations.

🔬 Research Interests

  • Federated Learning Security & Robustness
  • Data Poisoning & Backdoor Attacks
  • Membership Inference & Privacy Leakage
  • Privacy-Preserving ML (Differential Privacy, Secure Aggregation)
  • Distributed & Edge Machine Learning
  • Evaluation of ML Systems under Data Heterogeneity

🎓 Education

Sungkyunkwan University (SKKU) — Suwon, South Korea

M.S. in Computer Science and Engineering (2024 – Present)

  • STEM Scholarship Recipient
  • Research Track: Machine Learning Security & Federated Learning
  • Graduate Teaching Assistant (Labs & Exercises)

University of Oran 1 Ahmed Ben Bella — Oran, Algeria

M.S. in Computer Science (2018 – 2020)

  • Final Project: Design and Deployment of a FIWARE-based IoT Platform
  • Focus on smart systems, cloud integration, and real-time data pipelines

UHBC University — Chlef, Algeria

B.S. in Computer Systems (2015 – 2018)

  • Final Project: E-commerce and Online Auction Platform
  • Strong foundations in algorithms, databases, and software engineering

🧪 Research & Project Experience

  • Federated Learning Experiments
    • Designed and ran large-scale FL experiments under non-IID client distributions
    • Evaluated robustness across CNNs and lightweight architectures
    • Analyzed trade-offs between accuracy, privacy, and attack resilience
  • Privacy & Security Analysis
    • Studied membership inference attacks under temporal and system-level settings
    • Benchmarked defenses including sampling, perturbation, and data augmentation
    • Focus on realistic threat models, not worst-case assumptions only
  • WaterMed4.0 (PRIMA Project)
    • International collaboration on smart irrigation systems
    • Worked on IoT data ingestion, context management, and analytics pipelines
    • Collaborated with research teams in Spain and Turkey
  • Telecom Internship – Access Telecom (Algeria)
    • Cellular network performance analysis
    • Practical exposure to measurement tools, KPIs, and real operator data

🛠️ Technical Skills

Programming & ML

  • Python, C, PHP
  • PyTorch, Federated Learning frameworks
  • Experimental design & evaluation pipelines

Systems & Platforms

  • Linux, Docker
  • IoT Platforms (FIWARE, AWS IoT)
  • MQTT, LoRaWAN, NGSI-LD, JSON-LD

Research Tools

  • Reproducible experimentation
  • Performance profiling & benchmarking
  • Data analysis & visualization

🌍 Languages

  • Arabic — Native
  • French — Fluent
  • English — Fluent

🔗 Links & Contact

  • 📄 CV: (add link)
  • 💼 LinkedIn: (add link)
  • 💻 GitHub: (add link)
  • ✉️ Email: (add academic email)

I am actively interested in research collaborations, PhD opportunities, and ML/AI engineering roles focused on trustworthy and privacy-aware AI systems.