Catching clouds: Simultaneous optimization of the parameters of biological agent plumes using Dirichlet processes to best estimate infection source location

摘要

We describe a stochastic method using Dirichlet processes to derive mixture models that allow the numerical description of outbreaks of diseases with multiple sources. We show that existing disease models may be extended using this method and how this may be used in a practical context to support the simulated response to a mass casualty public health emergency.