论文部分内容阅读
目的应用时间序列模型对象山县麻疹疫情进行分析和预测。方法采用描述流行病学方法分析象山县2002—2013年麻疹疫情监测数据,并对该序列资料进行平稳化、定阶并估计参数;建立时间序列模型,以2014年发病资料作为模型预测效果的考核样本来评价模型可靠性。结果象山县麻疹发病率自2008年明显上升后,一直维持较低水平,但有小幅波动,且疫情按月分布没有明显集中性。对麻疹发病数序列建立ARMA模型,并应用该模型预测已知的2014年1—4月份麻疹发病数,模型预测值与实际值基本一致,且实际值均处于预测值95%的可信区间内。结论 ARMA模型对象山县麻疹发病情况的拟合结果满意,预测效果良好,可为该县麻疹的预测预警提供依据。
Objective To analyze and forecast the measles epidemic in Shanxian County using time series model. Methods Descriptive epidemiological methods were used to analyze the surveillance data of measles epidemics in Xiangshan County from 2002 to 2013, and the data of the measles epidemics were normalized, ranked and the parameters were estimated. The time series model was established to assess the predictive effect of the disease in 2014 Sample to evaluate model reliability. Results The incidence of measles in Xiangshan County has been kept at a low level since its obvious increase in 2008, but fluctuated slightly. There was no obvious concentration of measles on a monthly basis. ARMA model was established for the number of measles incidence and the model was used to predict the number of measles cases from January to April 2014. The model predicted value was basically consistent with the actual value and the actual values were within the 95% confidence interval of the predicted value . Conclusion The results of fitting the incidence of measles in ARMA model are satisfactory and the prediction results are good, which may provide evidence for the prediction and warning of measles in this county.