预测传染病的的英文论文.pdfVIP

Models, Prediction, and Estimation of Outbreaks of Infectious Disease Peter J. Costa, James P. Dunyak, Mojdeh Mohtashemi The MITRE Corporation pjcosta@mitre.org Abstract 1.1 Notation Conventional SEIR (Susceptible–Exposed–Infectious– S = S(t) = number of people in the population susceptible to Recovered) models have been utilized by numerous the disease at time t researchers to study and predict disease outbreak. By E = E(t) = number of people in the population exposed/ combining the predictive nature of such mathematical infected by the disease at time t models along with the measured occurrences of disease, a I = I(t) = number of people in the population who are more robust estimate of disease progression can be made. infectious at time t The Kalman filter is the method designed to incorporate R = R(t) = number of people in the population who have model prediction and measurement correction. recovered from the disease at time t Consequently, we produce an SEIR model which governs There are a number of parameters which will need to be the short term behaviour of an epidemic outbreak. The either modeled or estimated from the data. It is assumed mathematical structure for an associated Kalman filter is that these parameters are time invariant though more developed and estimates of a simulated outbreak are sophisticated efforts and information could produce time– provided varying models. A description of these parameters is listed below. 1. Introduction 1.2 Parameters Mathematical models have been used to study the outbreak of a number of infectious diseases [1, 2, 6]. In β = probability of disease t