Information Theoretic Adaptive Tracking of Epidemics in Complex Networks

TL;DR

This paper introduces an information theoretic adaptive framework for tracking epidemics in complex networks, leveraging dynamic Bayesian inference to improve monitoring and understanding of phase transitions.

Contribution

It presents a novel adaptive sampling method based on information theory and establishes conditions for phase transitions in network monitoring, improving epidemic threshold modeling.

Findings

Framework effectively tracks epidemic states in complex networks.

Identifies conditions for observation-dependent phase transitions.

Provides insights into network response to interventions.

Abstract

Adaptively monitoring the states of nodes in a large complex network is of interest in domains such as national security, public health, and energy grid management. Here, we present an information theoretic adaptive tracking and sampling framework that recursively selects measurements using the feedback from performing inference on a dynamic Bayesian Network. We also present conditions for the existence of a network specific, observation dependent, phase transition in the updated posterior of hidden node states resulting from actively monitoring the network. Since traditional epidemic thresholds are derived using observation independent Markov chains, the threshold of the posterior should more accurately model the true phase transition of a network. The adaptive tracking framework and epidemic threshold should provide insight into modeling the dynamic response of the updated posterior to active intervention and control policies while monitoring modern complex networks.