A Framework for Multi-source Privacy Preserving Epidemic Analysis

2025-06-27

AI Summary

This study addresses the challenge of collaborative modeling over heterogeneous, multi-source epidemiological data—some containing sensitive information and others protected via differential privacy. We propose the first privacy-preserving analytical framework integrating differential privacy, deep learning, and mechanistic infectious disease dynamics. Methodologically, we innovatively introduce differentially private synthetic financial data as auxiliary features and design a unified architecture capable of jointly modeling both raw and privacy-protected data, enabling end-to-end fusion and dynamic mechanism learning under rigorous privacy guarantees. Experiments demonstrate that, under privacy budgets ε = 1–5, our framework reduces epidemic prediction error by 12.7%–23.4% and accurately infers transmission parameters (R₀ estimation error < 8.5%), significantly outperforming baseline models. Our core contribution is the first cross-domain transfer of differentially private synthetic data to epidemiological modeling, establishing a novel paradigm that jointly optimizes privacy preservation, predictive accuracy, and mechanistic interpretability.

Technology Category

• Machine Learning: Privacy
• Natural Language Processing: Ethics — Bias, Fairness, Transparency & Privacy
• Search and Optimization: Learning to Search

Application Category

• Security and Privacy: Data transparency and provenance
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

It is now well understood that diverse datasets provide a lot of value in key epidemiology and public health analyses, such as forecasting and nowcasting, development of epidemic models, evaluation and design of interventions and resource allocation. Some of these datasets are often sensitive, and need adequate privacy protections. There are many models of privacy, but Differential Privacy (DP) has become a de facto standard because of its strong guarantees, without making models about adversaries. In this paper, we develop a framework the integrates deep learning and epidemic models to simultaneously perform epidemic forecasting and learning a mechanistic model of epidemic spread, while incorporating multiple datasets for these analyses, including some with DP guarantees. We demonstrate our framework using a realistic but synthetic financial dataset with DP; such a dataset has not been used in such epidemic analyses. We show that this dataset provides significant value in forecasting and learning an epidemic model, even when used with DP guarantees.

Problem

• Integrate multi-source data for epidemic analysis with privacy
• Combine deep learning and epidemic models for forecasting
• Apply differential privacy to sensitive epidemic datasets

Innovation

• Integrates deep learning with epidemic models
• Incorporates multi-source datasets with privacy
• Uses Differential Privacy for sensitive data