论文部分内容阅读
目的:应用随机时间序列分析法预测痢疾发病率。方法:应用SPSS11.5软件对某市2001～2007年痢疾逐月发病率进行ARIMA模型建模拟合,用所得到的模型对2008年各月痢疾发病率进行预测。结果:ARIMA(1,0,0)×(1,1,0)12模型很好地拟合了既往时间段上的痢疾发病率序列,对2008年各月发病率的预测值符合实际发病率变动趋势,且实际发病率均在95%可信限内。实际值与预测值的全年误差为13.02%,1～6月份误差为4.91%。结论:应用随机时间序列分析法对痢疾的发病率进行短期预测能够收到很好的效果,为痢疾的防控提供科学有效的依据。
Objective: To predict the incidence of dysentery by stochastic time series analysis. Methods: SPSS 11.5 software was used to simulate the monthly incidence of dysentery from 2001 to 2007 in ARIMA model. The incidence rate of dysentery in each month in 2008 was predicted by the obtained model. Results: The ARIMA (1,0,0) × (1,1,0) 12 model fitted well the past incidence of dysentery in the previous time period. The predicted incidence in each month in 2008 was in line with the actual incidence Trend of change, and the actual incidence are within 95% confidence limits. The annual error between the actual value and the predicted value was 13.02%, and the error from January to June was 4.91%. Conclusion: The short-term prediction of the incidence of dysentery using random time series analysis can receive good results and provide a scientific and effective basis for the prevention and control of dysentery.