Agent-based models for epidemic dynamics
Marcelo F C Gomes
Instituto de Física, Universidade Federal do Rio Grande do S
Instituto Balseiro and Centro Atómico Bariloche
Instituto de Física, Universidade Federal do Rio Grande do Sul
Last modified: June 26, 2007
There are many ways of modeling epidemics, the must common ones being mean field - with the use of differential equations mainly -, which focus on the transmission rate and period of infection; complex networks, focusing on the role played by the network of interaction between the subjects in the epidemic dynamics, relating the parameters which characterizes the network with the epidemic threshold; and, most recently, agent-based models, where one tries to understand what is the influence of agents’ mobility on the propagation of a given disease. The present work is on the last line of approach, presenting several mobility scenarios and comparing the results amongst them and with the previous two models, using different kinds of epidemiological structures - SIS and SIRS. We show that for low densities, we have the same qualitative behavior as the one presented for the mean field, despite mixing the system or not. We also show that the system may present a steady state that has oscillations, a result presented on networks.