Research¶

Our laboratory focuses on advanced research in Federated Learning (FL), privacy-preserving machine learning, and distributed optimization. We develop novel methodologies and systems to enable collaborative machine learning across decentralized devices or organizations without sharing raw data.

Key Research Directions¶

• Privacy-Preserving Federated Learning: Designing secure aggregation protocols, differential privacy techniques, and cryptographic methods to protect user privacy.
• Efficient and Scalable Optimization: Developing fast-converging algorithms to reduce communication overhead and handle heterogeneous data distribution (Non-IID).
• Incentive Mechanisms and Fairness: Formulating game-theoretic approaches to encourage client participation and ensure fair contribution reward distribution.
• Robustness and Security in FL: Defending against malicious clients, poisoning attacks, and model inversion attacks.