Stochastic models of HIV epidemic in homosexual populations - the effects of mixing patterns (Q1197706)

The paper compares a number of multi-stage Markovian models of HIV transmission, in discrete and continuous time, with the corresponding deterministic analogues. The comparison is achieved by means of computer simulation, so that the results take the form of qualitative tendencies observed for various choices of the parameters, including the mixing pattern. In all the figures depicted, the deterministic curve lays for most of its evolution between the extreme values found in the simulations. Certain systematic differences between stochastic and deterministic, analogous to those known for more simple models, were also noted.