To Discuss The Methods Of Predictive Models For Tumor Hospital Outpatient And Inpatient Amount

Objective:To compare the predicting effect of the number of inpatient and outpatient Amount in tumor hospital by dynamic series,linear regression,Logarithmic regression,Quadratic regression,Cubic regression,ARIMA models and provide a fitted model for hospital managers.Methods:The different statistical models were applied to predict the number of inpatient and outpatient in tumor hospital and compared their absolute errors,relative errors and error sum of square.Result:the relative absolute errors of ARIMA models in the prediction of the tumor hospital outpatient Amount,the relative absolute errors of ARIMA models was smallest 11.29%,and then cubic 15.28%, Quadratic 15.43%,linear 15.60%and dynamic series 17.38%,the Logarithmic regression were larger20.17%.In predict the number of inpatient in tumor hospital was smallest 9.35%,and then Quadratic and linear 12.52%,cubic 12.53%and dynamic series 15.91%, the Logarithmic regression were larger 17.96%.Conclusions- ARIMA model is one of the best models in predict the number of outpatient and inpatient Amount in tumor hospital,this model can provide more efficient massage about hospital development.