Optimization of the Choice of Individuals to Be Immunized Through the Genetic Algorithm in the SIR Model

摘要

Choosing which part of a population to immunize is an important and challenging task when fighting epidemics. In this paper we present an optimization methodology to assist the selection of a group of individuals for vaccination in order to restrain the spread of an epidemic. The proposed methodology is to build over the SIR (Susceptible/Infected/Recovered) epidemiological model combined to a genetic algorithm. The results obtained by the application of the methodology to a set of individuals modeled as a complex network show that the immunization of individuals chosen by the implemented genetic algorithm causes a significant reduction in the number of infected ones during the epidemic when compared to the vaccination of individuals based on a traditionally studied topological property, namely, the PageRank of individuals. This suggests that the proposed methodology has a high potential to be applied in real world contexts, where the number of vaccines is reduced or there are limited resources.