论文部分内容阅读
目的用SARIMA模型对龙岩市手足口病(HFMD)发病进行预测,为防控提供参考。方法用Eviews 5.0软件对《疾病监测信息报告管理系统》中龙岩市2008年5月至2014年12月HFMD的月发病例数进行SARIMA模型拟合与分析。结果 2008年5月至2014年12月,HFMD发病数呈周期性波动,通过自然对数和1次1阶非季节差分,拟合的相对最优模型为SARIMA(2,1,3)(1,1,1)12,表达式为(1-1.16L+0.77L2)(1+0.33L12)(1-L)(1-L12)log(yt)=(1-0.86L+0.32L3)(1-0.90L12)εt。预测结果吻合较好,平均相对误差为10.2%。结论 SARIMA模型可对龙岩市手足口病进行较准确预测,对预警有实用价值。
Objective To predict the incidence of hand-foot-mouth disease (HFMD) in Longyan by SARIMA model and provide reference for prevention and control. Methods Using Eviews 5.0 software, SARIMA model fitting and analysis of monthly incidence of HFMD in Longyan from May 2008 to December 2014 in “Disease Surveillance Information Reporting Management System” was performed. Results From May 2008 to December 2014, the incidence of HFMD fluctuated periodically. The relative optimal model fitted by SARIMA (2,1,3) (1 , 1, 1) 12 and the expression is (1-1.16L + 0.77L2) (1 + 0.33L12) (1-L) (1-L12) log (yt) = (1-0.86L + 0.32L3) 1-0.90L12) εt. The predicted results agree well with the average relative error of 10.2%. Conclusion The SARIMA model can predict hand-foot-mouth disease in Longyan City more accurately and has practical value for early warning.