运用哑变量技术 提高预测门急诊量的准确性

Use dummy variable to improve accuracy of forecasting numbers of outpatient

目的使用时间序列模型,对医院门急诊就诊人数进行预测。方法在线性回归方程中引用哑变量,修正周期性变化,降低预测的误差。结果使用Excel软件数据分析功能,得出含有哑变量的多元线性回归方程。结论当一个时间序列数据呈现周期性变化时,可以引入哑变量,并根据情况确定哑变量个数来反映这种周期性变化,这样有助于提高预测数据的准确性。

Objective To forecast numbers of outpatient by using time-series model. Methods Dummy variable is introduced into the linear regression equation to correct cyclical variation and reduce error of forecasting. Results Muhiple linear regression equation with dummy variable by using Microsoft Excel was obtained. Conclusion If data of time-series model emerges cyclical variation, it＇s proper to adopt dummy variable and determine the numbers to reflect the cyclical variation according to the situation, which can help improve the accuracy of forecasting.