多元统计方法在中医诊断研究中的应用

针对中医证的多维性、非线性和复杂性的特点,文章以多元统计学方法为切入点,以logistic回归分析、聚类分析、判别分析、主成分典型相关分析与因子分析等常见的统计方法为例,对其在中医诊断学中的应用进行总结、分析。同时,结合中医传统理论及其数理思想和方法,发现其在中医辨证中有一定的适用性,也存在一定的问题。

Application of multivariate statistical method in the diagnosis of traditional Chinese medicine

According to the characteristics of multi-dimension,non-linearity and complexity of traditional Chinese medicine (TCM) syndrome,this paper is based on the multivariate statistical methods,and takes logistic regression analysis,cluster analysis,discriminant analysis,typical correlation analysis,principal component analysis and other common statistical methods as examples to summarize and analyze the application in TCM diagnostics. At the same time,combined with the traditional theory of TCM and its mathematical thoughts and methods,it is found that it has certain applicability and problems in TCM syndrome differentiation.