ARIMA乘积季节模型在重庆市流行性乙型脑炎预测中的应用

发布时间：2018-02-11 13:38

本文关键词： 流行性乙型脑炎自回归求和移动平均乘积季节模型预测　出处：《重庆医科大学学报》2017年10期 　论文类型：期刊论文

【摘要】：目的:应用自回归求和移动平均(autoregressive integrated moving average,ARIMA)乘积季节模型预测重庆市流行性乙型脑炎(简称乙脑)发病数。方法:利用R软件对重庆市2006年1月到2015年6月乙脑报告病例数进行ARIMA模型建模拟合,选择预测模型进行相互比较确定最优模型。用2015年7至12月实际报告病例数与拟合值的比较来评价模型的预测效果,并对2016至2017年重庆市乙脑报告发病数进行预测。结果:重庆市乙脑发病人数呈逐年下降趋势,报告病例具有明显季节分布特征,ARIMA(0,0,1)×(1,1,2)12模型很好地拟合了时间序列,该模型赤池信息量准则(Akaike information criterion,AIC)、许瓦兹贝叶斯准则(Schwarz Bayesian criterion,SBC)值均最小且预测值与实际值的平均相对误差为0.12,平均绝对百分比误差为7.81%。进一步用该模型预测重庆市2016至2017年乙脑病例数分别为35例和32例,发病高峰仍是7至8月。结论:利用ARIMA乘积季节模型对乙脑发病数拟合较好,短期预测结果良好;与2015年比较,预测2016至2017年乙脑报告发病数略微减少。
[Abstract]:Objective: to predict the incidence of Japanese encephalitis (Japanese encephalitis) in Chongqing by using autoregressive integrated moving autoregressive integrated moving autoregressive autoregressive seasonal model of autoregressive autoregressive and autoregressive integrated moving Arima. Methods: using R software to report the cases of Japanese encephalitis B from January 2006 to June 2015 in Chongqing city. The number of cases reported was fitted with ARIMA model. The prediction model was selected to compare with each other to determine the optimal model. The prediction effect of the model was evaluated by comparing the actual reported cases with the fitting value from 2015 to December. The number of reported cases of encephalitis B in Chongqing from 2016 to 2017 was predicted. Results: the incidence of je in Chongqing showed a decreasing trend year by year, and the reported cases had obvious seasonal distribution characteristics. The information quantity criterion of the red pool is Akaike information criterion and the Schwarz Bayesian criterion is the smallest. The average relative error between the predicted value and the actual value is 0.12, and the average absolute percentage error is 7.81. Further, the model is used to predict Chongqing from 2016 to 2016. In 2017, the number of cases of encephalitis B was 35 and 32, respectively. Conclusion: the seasonal model of ARIMA product fits the incidence of je well, and the short-term prediction results are good, and compared with 2015, the number of reported cases of je from 2016 to 2017 is slightly reduced.
【作者单位】：重庆市疾病预防控制中心;
【基金】：重庆市卫生计生委科研面上资助项目(编号:2015MSXM091)
【分类号】：R512.32

【相似文献】