基于依赖模式集挖掘组方药物多维配伍规律
Mining the Compatibility Law of Multidimensional Medicines Based onDependence Mode Sets
作者：乔少杰(四川大学 计算机学院，四川 成都 610065)；唐常杰(四川大学 计算机学院，四川 成都 610065)；韩楠(成都中医药大学 基础医学院，四川 成都 610075)
Author：(School of Computer Sci., Sichuan Univ., Chengdu 610065, China)；(School of Computer Sci., Sichuan Univ., Chengdu 610065, China)；(School of Chinese Basic Medicine, Chengdu Univ. of T. C. M., Chengdu 610075, China)
收稿日期：2006-06-26 年卷（期）页码：2007,39(4):134-138
期刊名称：工程科学与技术
Journal Name：Advanced Engineering Sciences
关键字：数据挖掘；配伍规律；依赖性判定；差异函数
Key words：data mining; compatibility law; dependence evaluation; difference function
基金项目：国家自然科学基金资助项目(60473071)；高等学校博士学科点专项科研基金SRFDP(20020610007)；四川省青年软件创新工程(2005AA0816)
中文摘要
针对多维组方药物之间配伍规律挖掘问题，提出了依赖性判定函数和多维组方药物依赖性判定算法MMDE(Multidimensional Medicine Dependence Evaluation)，作为组方药物依赖性判定的依据，利用差异函数确定药物分布的离散度，提出了基于可调差异函数求解缺剂量药物用量的新方法。实验证明了新方法在挖掘组方药物间依赖关系上准确性较高，达到80%以上。
英文摘要
A concept of dependence evaluation function and MMDE (Multidimensional Medicine Dependence Evaluation) algorithm were proposed to determine the dependence relationship of multidimensional medicines aiming at mining the compatibility law of multidimensional medicines. The difference function was used to determine the divergence degree of the distribution of medicines, and a new method based on the modified difference function was proposed to compute the dosage of imperfect medicines based on the modified difference function. Experiments showed that the new method for mining the dependence relationship of medicines has accuracy greater than 80%.
【关闭】
