基于粗糙集与支持向量机的中医辨证数据挖掘方法研究

目的:采用粗糙集与支持向量机结合的数据挖掘方法进行中医临床肝硬化的辨证研究。方法:根据所收集的293例中医肝硬化患者临床的相关信息,采用粗糙集方法建立证候决策信息表,并通过计算提取与肝硬化各证型有密切关联的重要症状、体征。然后,利用这些提取的症状组合作为支持向量机的出入进行分类学习,从支持向量机分类结果得到相应的证候。结果:采用不同的输入指标所得到的辨证平均正确率均高于70%,尤其采用粗糙集约简后辨证的准确率为84.4%左右。结论:基于粗糙集与支持向量机的中医辨证具有比较高的可靠性,对于中医临床辨证研究工作具有一定的参考价值。

The Research for Data Mining Method of TCM Syndrome Identification Based on Rough Sets and SVM

Objective：To research Traditional Chinese Medicine（TCM） syndrome identification based on Rough sets and multi-class support vector machines（SVM）.Methods：Using the clinical data from 293 patients with posthepatitic cirrhosis,the information decision-making table was set up by Rough Sets method.Furthermore,it was found the characteristic symptom groups as related to the TCM Syndromes from the table.These groups were used as the inputs of SVM for classification and then the appropriate syndromes were gained according to the outputs of SVM.Results：All of the correctness of SVM classifications was above 70%.Especially,using Rough Sets,the correctness of SVM classification bring up as far as 84.4%.Conclusion：The result shows that the data mining method based on Rough Sets and SVM was reliable and referenced for TCM syndrome identification.