APP下载

基于ARIMA模型结合回归分析在产科工作量预测中的应用

2019-08-19李芷云常薇陈平陈洁芬

中国医学创新 2019年10期

李芷云 常薇 陈平 陈洁芬

【摘要】 目的:探讨ARIMA模型结合回归方程在产科工作量预测中的应用价值。方法:以产科门诊建卡数与分娩量数据为基础,运用SPSS19.0建立月建卡数的ARIMA模型、月建卡数与月分娩量的回归方程模型,采用实际数据验证模型,评价模型,选择精度较高的模型进行2018-2020年的产科工作量预测。结果:月建卡数的模型ARIMA(1,1,1)(0,1,1)12的拟合效果最优,用实际分娩量进行预测精度验证,预测分娩量与实际分娩量基本吻合,且实际值均在预测值可信区间范围内。符合拟合程度较高以及线性回归的显著性检验要求的回归方程为y=313.727+1.212x,其预测的平均误差为5.114%,具有较高的精确度,并对历史值的预测效果较好。采用ARIMA模型结合线性回归分析预测2018-2020年建卡人数分别为11 324、12 388、13 334,增长率依次为4.62%、9.40%、7.64%;分娩量预测值分别为12 936、13 554、14 369,增长率依次为9.61%、4.78%、6.01%。结论:ARIMA模型结合回归分析具有较高的预测精度,可较好地拟合产科工作量的演变趋势,为新生育政策下产科管理提供決策依据。

【关键词】 ARIMA模型; 回归分析; 建卡数; 分娩量; 预测

【Abstract】 Objective:To explore application of ARIMA model combined with regression analysis in prediction of obstetric workload.Method:Based on the number of health care cards in obstetric outpatient and delivery volume,the ARIMA model of the number of health care cards per month,the regression equation model of the number of health care cards per month and the delivery volume per month were established by SPSS 19.0.The actual data were used to validate the model,and the evaluation model was selected to predict the obstetric workload from 2018 to 2020. Result:The model ARIMA(1,1,1)(0,1,1)12 has the best fitting effect.The prediction accuracy was verified by the actual delivery volume.The predicted delivery volume basically coincides with the actual delivery volume,and the actual value was within the confidence interval of the predicted value.The regression equation which accords with the higher fitting degree and the significance test of linear regression is y=313.727+1.212x.The average error of prediction was 5.114%.It had higher accuracy and better prediction effect for historical values.ARIMA model combined with linear regression analysis was used to predict the number of health care cards from 2018 to 2020 were respectively 11 324,12 388,13 334,the growth rate were respectively 4.62%,9.40% and 7.64%;the predicted value of delivery volume were respectively 12 936,13 554,14 369,the growth rate were respectively 9.61%,4.78% and 6.01%.Conclusion:ARIMA model combined with regression analysis has a high prediction accuracy,can better fit the evolution trend of obstetric workload,provide decision-making basis for obstetric management under the new birth policy.

【Key words】 ARIMA model; Regression analysis; Number of health care cards; Delivery volume;Forecast

First-authors address:Wuhan University of Science and Technology School of Medicine,Wuhan 430081,China

doi:10.3969/j.issn.1674-4985.2019.10.041

我国自2014年以来先后推行了“单独二孩”和“全面二孩”政策[1],产科工作量近几年呈现上升趋势,直接冲击助产机构的服务能力,因此对产科工作量进行较为准确预测显得十分重要和迫切。国内产科工作量预测分析一般用分娩量或活产数[2-3],但考虑相关因素对分娩量影响的时间序列应用方法不多。分娩量的需求增长受到生育意愿、助产机构技术水平等众多因素影响,而基于孕……

登录APP查看全文