Stochastic Epidemic Modeling for the H1N1 Influenza in Korea

Abstract

We present computational methods of the stochastic epidemic models such as stochastic simulation algorithm(SSA) and moment closure method(MCM) and apply the methods to an H1N1 influenza model. Computational efficiency is an important issue for stochastic modeling when one simulate the model with different situations and parameters. In general, the exact stochastic simulation algorithm(SSA) requires intensive and expensive computations especially when the infection process is fast and there is a large number of population . In that case, the moment closure method(MCM) is a good alternative way of computation for stochastic modeling. A MCM is useful approximation method to find the moments by truncating the higher order from the ODE system of moments. There have been some moment closure methods described in [15,19–21]. We compare the results between deterministic and stochastic models in terms of the spread of the influenza. We show that there is a considerable discrepancy between the results of stochastic and deterministic models especially when a small number of either infective or susceptible individuals is initially.