不同数据分析方法用于出血性中风证候分析的对比研究

目的:探讨不同数据分析方法在出血性中风证候分析中的效果差异.方法:基于临床采集的499例出血性中风患者中医四诊信息数据,比较因子分析、聚类分析、Logistic回归分析、Fisher判别分析在证候分类及归纳、挖掘证候潜在规律的差异.结果:因子分析、聚类分析对出血性中风的证候分类较为模糊,公因子和聚类组中见多个证候要素夹...

Comparative Study of Different Data Analysis Methods Used in Syndrome Analysis of Hemorrhagic Stroke

Objective:To discuss the effect difference of various data analysis methods used in the analysis of hemorrhagic stroke syndrome.Methods:Based on the TCM four diagnostic information of 499 cases with hemorrhagic stroke,compare the differences between factor analysis,cluster analysis,Logistic regression,and Fisher discriminant analysis in the classification and induction of syndromes,and explore the differences in the Syndrome classification and induction,and excavation its underlying patterns.Results:Factor analysis and cluster analysis are vague in the classification of hemorrhagic stroke syndromes,and multiple syndrome elements mixed in the common factor and clustering combination.Logistic regression and Fisher discriminant analysis performed better in some syndromes,which were affected by a variety of factors such as prior factors.Conclusion:In TCM syndrome analysis,various data analysis methods have their advantages.Sole data analysis cannot meet different needs in syndrome analysis at the same time.It should be used in combination or introducing with new data analysis methods.