Antiga Prinyanilakul. Modeling spatial spread of epidemics using cellular automata. Master's Degree(Applied Mathematics). King Mongkut's University of Technology North Bangkok. Central Library. : King Mongkut's University of Technology North Bangkok, 2010.
Modeling spatial spread of epidemics using cellular automata
Abstract:
The spatial spread of epidemics can be studied by partial differential equations
(PDE), cellular automata (CA) or other methods. PDE models are suitable when the
spread of the infection is through local interactions. CA models are suitable when the
infection is spread either through local or non-local interactions. Non-local
interactions occur, for example, if infected individuals can travel from one location to
another. In this thesis, we consider models for two types of epidemics: susceptibleinfectious-
recovered (SIR) and susceptible-infectious-susceptible (SIS). For SIR-type
epidemics, we first summarize the properties of differential equation, discrete-time
and cellular automata models. Then, we start from two different partial differential
equation models and use a finite difference method to discretize the models. We then
use the discretized models to create CA models. Given a PDE model, a cellular
automata model can always be created by using finite differences to approximate the
derivatives. We also examine CA models created directly for both SIR-type and SIStype
epidemics. We develop Matlab computer programs to simulate the behavior of
all models and use the programs to obtain results for selected parameter values. For
SIR-type epidemics the qualitative behavior of solutions is similar for the PDE and
CA models, although there are differences in detail. For the SIS-type models, the
qualitative behavior of the solutions for the populations in individual cells of the CA
is very sensitive to parameter values. However, the qualitative behavior of the
populations in the complete CA lattice converges smoothly to equilibrium values.