江苏省2006-2014年城乡未成年人跌倒病例的时间序列预测分析

The fall injury cases of urban and rural areas for minors in Jiangsu Province:a time-series prediction and analysis,2006-2014

目的建立江苏省城乡地区未成年人跌倒发生情况的预测模型,并对城乡地区跌倒的发生情况进行预测分析,为未成年人跌倒的预防和控制提供参考依据。方法收集2006-2014年江苏省城乡未成年人每月跌倒的病例数并进行整理,以2006-2013年的数据作为拟合样本,以2014年的数据作为预测样本;采用统计软件SPSS18.0建立季节性自回归移动平均混合模型(seasonal autoregressive integrated moving average model,SARIMA model)并进行预测分析。结果通过模型构建最终拟合的模型城乡分别为SARIMA(0,1,1)(0,1,1)12和SARIMA(0,1,1)(1,1,0)12。模型诊断显示模型残差序列均为白噪音序列,且Box-Ljung检验所有Q统计量均无统计学意义,说明所建模型拟合度较好。2014年城市预测值与观测值之间平均相对误差为11.89%,农村相应的平均相对误差为2.10%;两地预测结果与观测结果基本一致。结论 SARIMA模型可对江苏省未成年人跌倒发生情况进行较准确的预测,可为及时、科学地预防和控制未成年人跌倒的发生提供可靠的数据支持。

Objective To establish a predictive model of fall injury of minors in urban and rural areas in Jiangsu Province, and predict the occurrence of fall injury in urban and rural areas in order to offer reference of preventing and controlling minors fall injury. Methods The minors fall eases of urban and rural areas in Jiangsu Province from 2006 to 2014 were collected and grouped by monthly occurrence. The data from 2006 to 2013 was used as the fitting samples, and the data for 2014 was used as the forecast samples. The seasonal autoregressive integrated moving average model （ SARIMA model） was established and analyzed by statistical software SPSS 18.0. Results Through the process of modeling, the ultimately fitting model was SARIMA（0,1,1 ） （0,1,1） 12 and SARIMA （0,1,1） （1,1,0） 12 for urban and rural areas, respectively. The model diagnosis showed the residual error was white noise, and the Box-Ljung test for the model showed Q statistic was not statistically significant, so the model fitted the data well. In 2014, the average relative error between predicted value and observed value was 11.89% in urban, the corresponding average relative error in rural was 2. 10%. The predicted results of the two places were nearly in accordance with the observations. Conclusions SARIMA model could forecast accurately for the falls of minors cases in Jiangsu Province, thus it can provide reliable data base in order to prevent and control cases of fall injury of minors more timely and scientifically.