论文部分内容阅读
目的分析北京市5岁以下儿童死亡率(U5MR)的长期变化趋势,探讨求和自回归滑动平均模型(ARIMA)预测U5MR的效果,并对2016-2020年北京市U5MR进行预测。方法基于北京市5岁以下儿童死亡监测网收集1992-2015年北京市5岁以下儿童死亡数据。以1992-2013年北京市U5MR为训练样本,拟合ARIMA模型。分别以1992-2013年和2014-2015年北京市U5MR为校验样本,对模型进行内外部验证。通过EViews 8.0和SPSS 19.0软件实现。结果 1992-2013年北京市U5MR呈下降趋势,22年间下降了81.50%。而2014和2015年略有回升,2年间上升了4.86%。构建最佳的死亡率预测模型为ARIMA(1,1,1):x_t=-0.445+1.509x_(t-1)-0.509x_(t-2)+α_t+0.999α_(t-1),R~2=0.982,MAPE=4.76%,AIC=2.15,BIC=2.30,满足白噪声。经内外部验证,实测值和预测值差异均无统计学意义(P_(内部)=0.401,P_(外部)=0.655)。利用此模型预测2016-2020年北京市U5MR,预测值分别为2.88‰、2.87‰、2.90‰、2.97‰和3.09‰。结论 1992-2013年北京市U5MR呈下降趋势,2014和2015年则略有回升。利用ARIMA(1,1,1)模型对北京市U5MR进行预测,效果较好。预测2016-2020年北京市U5MR呈现小幅上升趋势,提示应加大对北京市妇幼保健工作的投入和关注。
Objective To analyze the long-term trend of the mortality of children under 5 years old in Beijing, and to explore the effect of U5MR by the ARIMA model and to predict the U5MR in Beijing from 2016 to 2020. Methods Data on deaths of children under 5 years old in Beijing from 1992 to 2015 were collected based on the Beijing Children’s Death Monitoring Network for Children Under 5 Years. Taking Beijing U5MR from 1992 to 2013 as a training sample, the ARIMA model was fitted. Taking the U5MR of Beijing city from 1992 to 2013 and 2014 to 2015 as the calibration samples respectively, the model is verified internally and externally. Implemented with EViews 8.0 and SPSS 19.0 software. Results From 1992 to 2013, the U5MR in Beijing showed a downward trend, down 81.50% in 22 years. However, it rebounded slightly in 2014 and 2015, up by 4.86% in two years. The best model for predicting mortality was ARIMA (1,1,1): x_t = -0.445 + 1.509x_ (t-1) -0.509x_ (t-2) + α_t + 0.999α_ (t-1) ~ 2 = 0.982, MAPE = 4.76%, AIC = 2.15, BIC = 2.30, satisfying white noise. After internal and external validation, there was no significant difference between the measured value and the predicted value (P_ (internal) = 0.401, P_ (external) = 0.655). Using this model to predict Beijing U5MR from 2016 to 2020, the predicted values are 2.88 ‰, 2.87 ‰, 2.90 ‰, 2.97 ‰ and 3.09 ‰, respectively. Conclusion The U5MR in Beijing from 1992 to 2013 showed a downward trend, with a slight rebound in 2014 and 2015. Using ARIMA (1,1,1) model to predict U5MR in Beijing is better. It is predicted that there will be a slight upward trend of U5MR in Beijing from 2016 to 2020, suggesting that greater attention should be paid to maternal and child health care in Beijing.