初探用最小二乘法预测医院住院例数

Using the least square method to predict the number of hospital inpatients

目的:建立线性回归模型,预测本院2017-2020年住院例数。方法:采用最小二乘法建立线性模型,并对模型进行回归分析。结果:本院2017-2020年住院例数预测值分别为15 030、16 300、17 570和18 840,预测区间分别为12 230~17 830、13 500~19 100、14 770~20 370和16 040~21 640。结论:最小二乘法是从事物变化的因果关系出发来进行预测的一种方法,通过预测住院例数,能为医院的工作计划和决策提供理论依据,使卫生资源合理应用。

Objective:A linear regression model was established to predict the number of inpatients in the hospital for2008-2016years.Methods:The linear model was established by the least square method,and the regression analysis of the model was carried out.Results:The estimated number of inpatients in the hospital for2017-2020years was15030,16300,17570and18840,respectively,the prediction interval was12230～17830,13500～19100,14770～20370and16040～21640respectively.Conclusion:The least square method is a prediction method based on the causal relationship of things.By forecasting the number of hospitalized patients,it can provide a theoretical basis for the work plan and decision-making of hospitals,and make rational use of health resources.