数据挖掘方法在中医证候学中的应用

Investigate the Statistical Analysis Method for the Diagnosis of Syndromes in TCM

[目的]探讨中医证候诊断的统计学分析方法。[方法]采用决策树、贝叶斯网络、神经网络、多元统计、聚类、关联规则等多种数据挖掘(data mining,DM)的统计方法,从众多的中医病症数据中,寻找规律性,提取出事先未知而潜在有用的信息,归纳中医证候的辨证分型。[结果]可作为实现中医证候规范化的一种工具。[结论]符合中医病症数据特点的统计方法,在证候研究中越来越受到重视;联合应用DM的数种方法进行中医证候的探索可起到事半功倍的效果。

[Objective]To investigate the statistical analysis method for the diagnosis of syndromes in TCM.[Methods]Using decision tree,bayesian networks,neural networks,multivariate statistics,clustering,association rules,and other statistical methods of data mining,data from a host of TCM symptoms,seek regularity,to extract the unknown and potentially useful information in advance and syndrome differentiation of traditional Chinese medicine.[Results] It can implement type of syndrome standardization as a tool.[Conclusion]According with the characteristics of TCM symptoms data statistics,more and more shall be taken seriously in the study of syndrome;Combined use of several kinds of methods for data mining exploration of syndromes can have twice the result with half the effort.