摘要
目的 :探讨多元线性回归模型研究住院费用影响因素时存在的问题及原因 ,并提出相应的改进措施。方法 :选用循环的模式逐步建立影响其住院费用的适宜模型。结果 :由于线性假设不成立 ,多元线性回归建立的模型拟合效果不理想 ,此时应放松模型的线性约束。结论 :建立统计模型应对模型的有关假设进行验证 。
Objective:The problem of multiple linear regression model to analyse the influent factor of the cost of impatient and the reason which caused and how to improve on it was discussed.Methods:Circulatory mode was applied to conduct appropriate model.Results:Due to linear assumption is untenable,the effect of multiple linear regression model is not good.Here,the regression techniques to free from the restrictive assumption of classical regression would be applied.Conclusion:In order to assure of the accuracy of model,the regression diagnosis would be applied to test the assumption of model.
出处
《现代预防医学》
CAS
2002年第1期33-34,共2页
Modern Preventive Medicine
