传染病预测及模型选择研究进展

Research progress on prediction and model selection of infectious diseases

传染病是对人类社会存在威胁的全球公共卫生问题,传染病早期预警将大大降低传染病的社会经济危害。目前应用于预测传染病的数学模型和方法体系较为广泛的主要是传统方法、传播动力学模型以及多元统计。传统方法主要指回归方法、灰色模型、时间序列模型以及马尔科夫模型等,传播动力学模型的经典形式是SIS模型和SIR模型。多元统计如维度缩减、联合建模和向量累积等方法近年来也运用较多。模型研究对疾病的预测预警有重要意义。

Infectious disease is a global public health problem that threatens human society. Early warning of infectious diseases will reduce the social and economic harm of infectious diseases. The mathematical models and methods are applied to predict infectious diseases. Traditional methods, transmission dynamics model and multivariate statistical methods are widely used. Traditional method smainly refers to the regression method, grey model, time series model and markov model. The classic form of transmission dynamics model is the SIS model and SIR model. Multivariate statistic methods such as dimension reduction, joint modelling methods and vectorac cumulation methods are frequently used nowadays. Model research is of great significance to the prediction and early warning of disease.