多因子共线性的主成分logistic回归分析

Logistic Regression Based on Principal Component Analysis in Resolving the Co-linearity

目的 采用主成分改进的条件logistic解决原发性开角型青光眼病因研究中的共线性问题。方法 采用1∶3配比病例对照设计 ,利用多重线性回归中的共线性诊断技术进行共线性诊断 ,以主成分改进的条件logistic回归分析解决共线性。结果 主成分条件logistic回归能改善参数估计 ,显著降低标准误。 结论 主成分改进的条件logistic能很好地解决原发性开角型青光眼病因研究中的共线性问题。

Objective Adopting logistic regression model based on principal component analysis to resolve the co-linearity in the exploring of the risk factors of POAG.Methods Using 1∶3 matched case-control study,making the co-linearity diagnosis by related methodsand employing Logistic regression based on principal component analysis to resolve the problem.Results Logistic regression model based on principal component analysis can improve the estimating of parameters and decrease the standard errors significantly.Conclusion Logistic regression model based on principal component analysis can resolve the co-linearity in the exploring of the risk factors of POAG.