论文部分内容阅读
目的对医院出院人次进行预测分析,旨在为医院科学管理提供参考依据。方法对某院2006年1月-2014年12月的出院人数进行SARIMA模型的建模拟合,利用2015年1月-12月的出院人数对模型进行检验,然后利用所建立的模型预测2016年的出院人数。结果 SARIMA(2,1,1)(0,1,1)12(不含常数项)是该院出院人数的最优拟合预测模型,模型的R2=0.956,贝叶斯信息准则(BIC)为9.05;模型的平均相对误差绝对值为3.60%;2016年的出院人数预测值为22111。结论出院人数具有季节变动及长期增长趋势,SARIMA模型对于短期的医院出院人数预测具有较强的实用价值,可以为科学管理提供参考依据。
Objective To predict the hospital discharge number, which is intended to provide a reference for the hospital scientific management. Methods The SARIMA model was fitted to the number of discharged patients from January 2006 to December 2014 in a hospital. The model was tested by the number of discharged patients in January-December 2015 and then the model was used to predict 2016 The number of discharges. Results SARIMA (2,1,1) (0,1,1) 12 (excluding constant items) was the best fitted prediction model for the number of discharged patients in this hospital. The R2 = 0.956, Bayesian Information Criterion (BIC) Was 9.05. The absolute mean relative error of the model was 3.60%. The projected number of discharged patients in 2016 was 22111. Conclusion The number of discharged patients has a seasonal change and long-term growth trend. The SARIMA model has strong practical value for the short-term hospital discharge number prediction and can provide a reference for scientific management.