Bayesian Epidemic Detection in Multiple Populations

TL;DR

This paper introduces a Bayesian epidemic detection method that fuses spatial information from multiple populations, optimizing detection timing while considering costs and false alarms, demonstrated through synthetic examples.

Contribution

It presents a novel Bayesian framework for epidemic detection across multiple populations, incorporating spatial data fusion and cost-benefit analysis, with an efficient simulation-based solution.

Findings

Adaptive detection outperforms threshold-based strategies

The detection map visualizes system state and policy relationship

Simulation results validate the method's effectiveness

Abstract

Traditional epidemic detection algorithms make decisions using only local information. We propose a novel approach that explicitly models spatial information fusion from several metapopulations. Our method also takes into account cost-benefit considerations regarding the announcement of epidemic. We utilize a compartmental stochastic model within a Bayesian detection framework which leads to a dynamic optimization problem. The resulting adaptive, non-parametric detection strategy optimally balances detection delay vis-a-vis probability of false alarms. Taking advantage of the underlying state-space structure, we represent the stopping rule in terms of a detection map which visualizes the relationship between the multivariate system state and policy making. It also allows us to obtain an efficient simulation-based solution algorithm that is based on the Sequential Regression Monte Carlo (SRMC) approach of Gramacy and Ludkovski (SIFIN, 2015). We illustrate our results on synthetic examples and also quantify the advantages of our adaptive detection relative to conventional threshold-based strategies.