A complex network-based vaccination strategy for infectious diseases

Abstract

In the event of an outbreak, it is vital to stop the spread of infectious diseases as quickly as possible, and vaccination is the most important means of stopping the spread of infectious diseases. The research of infectious disease vaccination strategy aims to control the spread of infectious disease as effectively as possible under the circumstance of limited vaccine resources, so as to achieve the result of social group immunity. Most proposed solutions have been based on variations on infectious disease models, or on the structure of networks for vaccination strategies. In this study, the epidemic threshold is used as the optimization objective, which not only transforms the immunization problem into an optimization problem, but also can be applied to different infectious disease models and is more relevant for the study of infectious diseases relative to the network structure. Meanwhile, this work combines machine learning with evolutionary computation, which in turn improves the ability to search for big data problems and reduces computational costs. Community structure is presented to narrow the search based on complex networks. Therefore, we present a novel community-based targeted immunization framework(CTIF) to select nodes for immunization. The proposed CTIF is composed of three phases: community detection, candidate immunization node set generation and target immunizations node set selection. In terms of epidemic threshold optimization, experiments reveal that the suggested algorithm outperforms the baseline strategies.