基于信息熵的决策树在慢性胃炎中医辨证中的应用

Application of decision tree based on entropy in traditional Chinese medicine symptom analysis of chronic gastritis

目的 :探讨基于信息熵的决策树在慢性胃炎中医辨证分型中的应用。 方法 :采用 bootstrap方法对 4 0 6例样本进行扩增以满足数据挖掘对样本量的要求 ,采用基于信息熵的决策树 C4 .5算法建立中医辨证模型。结果 :决策树 C4 .5算法筛选出对中医辨证分型有意义的 2 6个因素并对其重要性进行排序 ;产生清楚易懂可用于分类的决策规则 ;建立辨证模型 ,模型分类符合率为 :训练集 83.6 0 % ,验证集 80 .6 7% ,测试集 81.2 5 % ;模型区分各类证型的灵敏度和特异度也较高。 结论 :决策树C4 .5算法建立的模型效果较好 。

Objective:To explore the application of decision tree based on entropy in traditional Chinese medicine (TCM) symptom analysis chronic gastritis. Methods: Bootstrap methods were used to multiply 406 cases for data mining, and models for TCM symptom were built using decision tree of C4.5.Results: Twenty-six important factors were selected and ranked according to the importance; readable diagnostic rules were produced and a model was built, of which the correctly classified rate of training set, validation set and test set were 83.60%, 80.67% and 81.25%,respectively. The sensitivity and specificity of the model to differentiate the TCM symptom was good. Conclusion: The model built by C4.5 is satisfactory and can be used for differentiating diagnosis of chronic gastritis with TCM symptom.