论文部分内容阅读
目的探讨求和自回归移动平均(ARIMA)模型在流行性腮腺炎发病预测中的应用,验证分析模型的可行性与适用性。方法对南京市2004年1月至2012年12月流行性腮腺炎发病率资料进行ARIMA模型拟合,用建立的模型对2013年1—12月发病率进行拟合检测,之后对2014年各月发病率进行预测评价。结果 2004—2013年流行性腮腺炎累计报告病例14 871例,年均发病率为21.78/10万,每年各月流行性腮腺炎发病率始终围绕在1.85/10万附近波动。建立ARIMA(1,0,0)(2,1,0)12模型为最优模型。模型残差序列为白噪声。除常数项外,模型各参数均有统计学意义。模型的平均绝对百分误差为29.63%,R2为0.76。用建立的模型拟合2013年1—12月发病率,均在95%可信区域内,符合实际发病率变动趋势,验证了该模型的可行性。用该模型对2014年流行性腮腺炎进行预测,年发病率为1.48/10万,发病高峰期在4、5、6月,月发病率分别为2.33/10万、2.72/10万、2.52/10万。结论 ARIMA模型可用于拟合流行性腮腺炎发病率在时间序列上的变化趋势,可进行动态分析和短期预测。
Objective To investigate the application of the ARIMA model in the prediction of the incidence of mumps and to verify the feasibility and applicability of the model. Methods The data of incidence of mumps in Nanjing from January 2004 to December 2012 were fitted by ARIMA model. The incidence of mumps was fitted with the established model from January to December in 2013, The incidence of predictive evaluation. Results A total of 14 871 cumulative reported cases of mumps were reported in 2004-2013, with an average annual incidence of 21.78 / 100 000. The incidence of mumps fluctuated around 1.85 / 100 000 in each month every year. Establishing ARIMA (1,0,0) (2,1,0) 12 model as the optimal model. The model residual sequence is white noise. In addition to the constant term, the model parameters were statistically significant. The average absolute percentage error of the model was 29.63% and R2 was 0.76. Using the established model to fit the incidence of January-December 2013, all of them were within the 95% confidence interval, which accorded with the changing trend of actual morbidity. The feasibility of the model was verified. The model was used to predict the epidemic mumps in 2014 with an annual incidence rate of 1.48 / lakh. The peak incidence was in April, May and June, with a monthly incidence of 2.33 / lakh, 2.72 / lakh and 2.52 / 100,000. Conclusions The ARIMA model can be used to fit the trend of mumps morbidity in time series. It can be used for dynamic analysis and short-term prediction.