ARIMA乘积季节模型在医院门急诊人次预测中的应用  

Application of the Model of Multiple Seasonal ARIMA in Forecasting the Hospital Outpatients' Visits

作　　者：范晓欣[1] 隋虹[1] 

机构地区：[1]哈尔滨医科大学公共卫生学院,黑龙江哈尔滨150081

出　　处：《中国医院管理》2015年第4期41-42,共2页Chinese Hospital Management

摘　　要：目的用ARIMA乘积季节模型(p,d,q)(P,D,Q)s预测门急诊人次,为医院管理策服务。方法数据源于某院HIS,选取2002年1月—2014年10月逐月门急诊人次数据,其中2002—2013年各月数据用于建立模型,2014年1—10月数据用于验证所建立的模型,然后用所建立模型预测2015年门急诊人次。用平均相对误差绝对值(MAPE)评价模型的预测精度。结果建立了ARIMA(2,1,1)(0,1,1)12模型,模型预测的MAPE为7.01%,2015年门急诊人次预测值为550 121。结论 ARI-MA乘积季节模型(p,d,q)(P,D,Q)s预测医院门急诊人次的效果理想,可以为医院管理者提供有价值的信息。Objective To apply a model of ARIMA （p,d,q）（P,D,Q）s to forecast the hospital’s visits to give service to managements and decisions in hospital. Method Collecting data from Jan. 2002 to Oct. 2014,in which the monthly data from the year 2002 to 2013 were used to establish the ARIMA model, and using the data from Jan. to Oct. in 2014 to test the model, then applying the built model to forecast the visits in 2015, and using the mean relative error absolute（MAPE）to evaluate the prediction’s precision. Result The model of ARIMA （2,1,1）（0,1,1）12 is built ,its MAPE of the prediction is 7.01% and the predicted hospital’s visits in 2015 is 550121. Conclusion The predicted result of the ARIMA （p,d,q）（P,D,Q）s in forecasting the hospital’s visits is good, which can supply the valuable information for managers in the hospital.

关 键 词：ARIMA乘积季节模型 医院门急诊人次 预测 

分 类 号：R197.323[医药卫生—卫生事业管理]