摘要
首先对现有的模糊关联规则挖掘算法中的支持度定义进行了改进,该支持度采用基于距离的定义方式.在此基础上,将决策树中目标属性的概念引入模糊关联规则挖掘中,提出了分类模糊关联规则挖掘模型;该模型在工业生产过程和现实生活中具有广泛的应用背景,而且较完全无指导的购物篮型关联规则挖掘问题更易于实现.最后,将这种分类模糊关联规则挖掘模型应用于芳烃抽提生产过程,通过对芳烃抽提生产过程的历史数据进行挖掘,找到了在芳烃生产过程中,当苯塔装置进料组分发生变化时,生产率降低、能耗增加的原因.
The existing definition of support was improved by adopting a new definition based on distance. Then classified fuzzy association rules mining model （CFARMM） was proposed. It takes association rule mining techniques applicable to classification tasks by introducing the concept of the target variable of decision tree. CFARMM is not only widely applicable in industrial production process and real life, but also easily realized as compared with shopping baskets association rules mining problem. Finally, CFARMM was applied to mine the real aromatic hydrocarbon extraction process. Through the data mining of historical process, the reasons that recovery rate decreased, and energy loss increased when the feed components varied were found.
出处
《化工学报》
EI
CAS
CSCD
北大核心
2005年第11期2137-2141,共5页
CIESC Journal
基金
国家杰出青年科学基金项目(60025308)
高等学校优秀青年教师教学和科研奖励基金项目资助.~~
关键词
数据挖掘
关联规则
支持度
芳烃抽提
data mining
association rules
support
aromatic hydrocarbon extraction
