SARS疫情传播的时间序列分析  

Analysis of Time Series of Transmission of SARS

作　　者：高世泽[1] 

机构地区：[1]重庆师范大学数学与计算机科学学院,重庆400047

出　　处：《重庆师范大学学报（自然科学版）》2004年第3期8-12,共5页Journal of Chongqing Normal University:Natural Science

摘　　要：利用时间序列分析的思想[1～3],对北京市2003年4～6月的累计确诊SARS病例进行研究,获得了日增确诊病例的变化趋势方程,并用自回归模型AR(20)来拟合传播过程,经方差分析知模型效果高度显著,将预测值与实际值比较,结果比较理想。由此推定每个SARS病人可以直接造成他人感染的期限平均在20天左右。In this paper the accumulative cases of SARS that have been made a definite diagnosis from April to June, 2003, in Beijing are investigated by analysis of time series. We give the trend equations of incremental cases of SARS every day. We also give a model AR(20) of Transmission of SARS. The model is notable at high level by analysis of variance. It is very good by comparing the predicted values with observed values. Therefore, we infer that the average time that a patient with SARS transmit other one is about 20 days.

关 键 词：SARS疫情 时间序列分析 自回归模型 趋势方程 传播过程 方差分析 

分 类 号：O211.61[理学—概率论与数理统计]