基于决策树的血瘀证诊断规则自动归纳方法

Method for Self-Extracting Diagnostic Rules of Blood Stasis Syndrome Based on Decision Tree

决策树是一种从数据中归纳规则的方法。利用决策树从290例血瘀证病例中自动地提取相应的诊断规则,得到决策树分类模型并归纳出5条血瘀证的诊断规则,对194例血瘀证病例测试结果为:阳性检测正确率、阴性检测正确率和检测正确率分别达到97.67%9、9.07%和98.45%。实验结果表明决策树能自动从中医病例中归纳诊断规则。

Decision tree is a method to induce rules from data. In this paper, decision tree was applied to self-extract diagnostic rules from 290 patient records related to blood stasis syndrome. Thus the decision tree classification model was obtained and five diagnostic rules were induced from the model. Applying these rules to classify 194 patient records,the results were 97.67% of sensitivity, 99.07% of specificity and 98.45% of accuracy. The experimental results show that the decision tree method is likely a promising method to self-extract diagnostic rules from patient records of Chinese medicine.