预测模型在门诊量分析与预测中的应用

目的分析门诊人次数据,探讨SPSS12.0软件中时间序列模型在门诊量数据分析中的应用方法。方法根据某医院1998—2005年门诊人次历史资料,研究数据的季节性特征,指数平滑模型、自回归模型以及ARIMA模型的应用情况,并对三模型进行比较。结果ARIMA模型的平均预测相对误差较小,预测效果最好。结论预测方法有很多种,并没有一种方法适用于各种情况,因此针对不同的情况选择合适的预测方法非常重要,ARIMA模型适用于医院门诊人次数据预测。

Application of prediction model to analyze and predict outpatient quantity

Objective To analyze outpatient quantity, discuss the methods and applications of time series model in SPSS 12.0. Methods Seasonal character, exponential smoothing model, autoregression model, ARIMA model were analyzed and the prediction effects of them were compared based on the outpatient quantity from Jun, 1998 to Dec, 2005. Results The average relative error of ARIMA model is less, which shows a satisfactory effectiveness. Conclusion No single model is the best one which suits all kinds of prediction. It is very important to select suitable methods. MRIMA model fits for outpatient quantity statistical prediction.