霍特双参数指数平滑预测模型在医院管理中应用的探讨  

Exploration of application of Holt's linear exponential smoothing in hospital management

作　　者：杨建南[1] 汪学军[1] 刘勇华[2] 何均华[3] 

机构地区：[1]成都铁路中心医院,四川省成都市610081 [2]成都市卫生局 [3]成都铁路局科学技术研究所

出　　处：《中国医院统计》2008年第3期193-195,共3页Chinese Journal of Hospital Statistics

基　　金：成都铁路局科技项目(CX0233)

摘　　要：目的探讨霍特双参数指数平滑预测模型在医院管理中的应用价值。方法建立霍特双参数指数平滑预测模型对成都市2006年住院人数进行定量预测,并对预测模型的预测精度进行评价。结果霍特双参数指数平滑预测模型预测住院人数的结果满意。通过计算机反复试算,可获得最佳的一组平滑常数(α,γ),使预测模型的预测误差最小。本例预测平均绝对百分误差(MAPE)3.4438%<10%,预测精度较高。结论霍特双参数指数平滑预测模型预测精度较高,计算简便且具有递推性质,在医院管理中有应用价值。Objective To explore the applied value of Holt＇s linear exponential smoothing. Methods The number of patients in hospitals during 2006 in Chengdu was calculated quantitatives by the means of Holt＇s linear exponential smoothing and the calculation accuracy was also evaluated. Results There was a satisfactory resulte to calculate the number of patients by the means of Holt＇s linear exponential smoothing. By counting repeatedly in computer, a group of exponential smoothing convention figure could be obtained which minimized the errors. The MAPE was 3. 443 8% in this case with high accuracy. Conclusion There was high accuracy of Holt＇s linear exponential smoothing. Because it was simple to calculate and could infered in sequential order, there was the applied value in hospital management.

关 键 词：霍特双参数指数平滑模型 预测精度 预测误差 住院人数 卫生服务需求 

分 类 号：R195.1[医药卫生—卫生统计学]