A global stochastic modeling framework to simulate and visualize epidemics.

摘要

This thesis proposes a framework to simulate and visualize the spread of an infectious disease in a population of a region such as a county. As real-world populations have a nonhomogeneous demographic and spatial distribution, this framework models the spread of an infectious disease based on population of and geographic distance between census blocks; social behavioral parameters for demographic groups. The population is stratified into demographic groups in individual census blocks using census data. Infection spread is modeled by means of local and global contacts generated between groups of population in census blocks. The strength and likelihood of the contacts are based on population, geographic distance and social behavioral parameters of the groups involved. The disease dynamics are represented on a geographic map of the region using a heat map representation, where the intensity of infection is mapped to a color scale.;This framework provides a tool for public health personnel and epidemiologists to run what-if analyses on disease spread in specific populations and plan for epidemic response. By the means of demographic stratification of population and incorporation of geographic distance and social behavioral parameters into the modeling of the outbreak, this framework takes into account non-homogeneity in demographic mix and spatial distribution of the population. Generation of contacts per population group instead of individuals contributes to lowering computational overhead. Heat map representation of the intensity of infection provides an intuitive way to visualize the disease dynamics.