PHED-FL – Public Health Emergency Dataset for Federated Learning

The PHED-FL dataset is a curated collection of public health records designed to support research in federated learning applications during public health emergencies. It includes anonymized, multi-institutional data relevant to outbreak detection, patient triage, and resource allocation. The dataset simulates real-world scenarios where data privacy and decentralization are critical, enabling the development and evaluation of federated learning models without compromising sensitive health information. PHED-FL includes structured health features such as patient demographics, symptoms, diagnoses, and temporal markers of health events, making it suitable for a variety of machine learning and public health analytics tasks.