Learning graph topology from metapopulation epidemic encoder-decoder

2026-03-02

AI Summary

This work addresses the challenge of jointly inferring transmission parameters and inter-subpopulation migration networks in metapopulation epidemic models when prior knowledge of migration is unavailable. Existing approaches struggle to simultaneously estimate both components without strong assumptions. To overcome this limitation, the authors propose two encoder–decoder–based deep learning architectures that, for the first time, enable joint inference of migration graph topology and epidemiological parameters directly from time-series infection data, without requiring any prior assumptions about either component. The incorporation of multi-pathogen data substantially enhances the accuracy of topological reconstruction. Extensive experiments on diverse synthetic and real-world migration networks demonstrate that the proposed methods consistently outperform existing techniques, confirming their effectiveness and robustness.

Technology Category

• Search and Optimization: Metareasoning and Metaheuristics
• Machine Learning: Deep Neural Architectures and Foundation Models
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Web Mining and Content Analysis: Models for Web evolution
• Social Networks and Social Media: Influence propagation, information diffusion, and the prediction on networks

Abstract

Metapopulation epidemic models are a valuable tool for studying large-scale outbreaks. With the limited availability of epidemic tracing data, it is challenging to infer the essential constituents of these models, namely, the epidemic parameters and the relevant mobility network between subpopulations. Either one of these constituents can be estimated while assuming the other; however, the problem of their joint inference has not yet been solved. Here, we propose two encoder-decoder deep learning architectures that infer metapopulation mobility graphs from time-series data, with and without the assumption of epidemic model parameters. Evaluation across diverse random and empirical mobility networks shows that the proposed approach outperforms the state-of-the-art topology inference. Further, we show that topology inference improves dramatically with data on additional pathogens. Our study establishes a robust framework for simultaneously inferring epidemic parameters and topology, addressing a persistent gap in modeling disease propagation.

Problem

• metapopulation epidemic
• graph topology inference
• epidemic parameters
• mobility network
• joint inference

Innovation

• graph topology inference
• metapopulation epidemic model
• encoder-decoder architecture
• joint parameter estimation
• deep learning