Privacy-Preserving Federated loT Learning for Smart Public Health Surveillance

Researchers aim to develop a learning framework for public health surveillance, enhancing early disease outbreak detection and targeted interventions while ensuring individual data privacy by addressing potential data leakage.

Funded by the CCI Hub

Project Investigators

• Principal Investigator (PI): Thang N. Dinh, Virginia Commonwealth University Department of Computer Science
• Co-PI: Chen Chen, University of Virginia Department of Computer Science

Rationale

As Al-driven systems become integral to public health surveillance for outbreak detection and epidemic control, they face critical challenges in preserving data privacy while enabling comprehensive analysis. 

Key areas include potential data leakage vulnerabilities in cross-client communication and evolving cyber threats to the  federated learning process.

Projected Outcomes

Researchers will develop a novel framework that leverages Al techniques to enhance the cybersecurity of federated learning, a cutting-edge Al paradigm that enables collaborative model training through decentralized computation. 

The framework will enhance the security and privacy of Al-based federated learning, enabling secure early detection of outbreaks and targeted interventions. 

The research will also demonstrate the power of Al in bolstering cybersecurity measures for critical public health applications.