Characterization of an Ising-like SI model

Abstract

An Ising-like system was used to model a computer network through which malware is spreading. The rate of spread of infection was estimated by fitting the infection curves with a standard logistic model and an exponential model. Graphical analysis shows that the exponential decay model fit the infection curves better than the logisitic model. Trends in the residuals suggest that current mathematical models for epidemics are not consistent with the system.