论文部分内容阅读
探讨了 COX模型中出现多因子共线性时的处理方法 :首先应用主成分分析法将多因子的信息综合 ,保留几个主成分 ,然后以主成分作为新的因子进行 COX模型拟合。结果显示 ,该方法在一定程度上可以避免因共线性导致模型在解释上造成的矛盾以及参数估计的不稳定性。
In this paper, the coprocessing method of multi-factor collinearity in COX model is discussed. Firstly, the principal component analysis is used to synthesize multi-factor information, retain several principal components, and then use the principal component as a new factor to fit COX model. The results show that this method can avoid some inconsistencies caused by the collinearity caused by collinearity and the instability of parameter estimation.