一种高效挖掘冠心病医案关联规则的仿真研究

研究中医冠心病医案,高效挖掘有益信息规则问题,由于中医医案数据量大、关联性强,针对传统的关联规则挖掘算法处理中医医案数据时存在效率低、收敛速度慢及漏报规则等问题,提出一种小生境技术和基因表达式编程相结合的挖掘关联规则的方法。通过惩罚函数设置支持度阈值,利用小生境技术执行小生境演化、融合算法,结合基因表达式编程算法操作简单、鲁棒性强的优势搜索强关联规则,有效避免了算法早熟,解决了规则冗余。针对治疗冠心病的中医医案进行了验证性实验,实验结果表明,改进算法在提取有效信息的效率上有较大的提高,挖掘结果对冠心病中医临床诊治具有一定的参考价值。

Simulation Research on Efficient Mining Association Rules in Clinical Data of Coronary Heart Disease

The problems that TCM clinical data exist in a great deal of data and high,and traditional methods of mining association rules have the probems of low efficiency,slow convergence and rules omission.The paper proposed a new combined method based on niche technology and gene expression programming.The method designs the penalty function to set threshold,uses niche technology to finish evolution and integration,and combines with the advantages of simple and robust of gene expression programming to search the strong association rules.The medical treatment records of coronary heart disease were verified by the experiments.Experimental results show that,compared with traditional association rules mining method,the combined algorithm performs better in terms of diversity of population and discovering more effective association rules.The mining result has reference value in TCM treatment of the coronary heart diseases.