利用抽样技术和元学习的分布式关联规则挖掘算法  

Distributed association rules mining algorithm by sampling and meta-learning

作　　者：李梅花[1] 王黎明[1] 许红涛[1] 

机构地区：[1]郑州大学信息工程学院,河南郑州450052

出　　处：《计算机应用》2006年第4期872-874,877,共4页journal of Computer Applications

基　　金：河南省自然科学基金资助项目(0211050110)

摘　　要：结合动态项集计数技术和抽样的思想,利用元学习策略来产生频繁项集,提出了一个不共享内存的分布式关联规则挖掘算法DASM;引进了相似度的概念,并用之提高了挖掘的精确度。理论分析以及在IBM数据生成器生成的数据集上的实验均表明,DASM算法具有较高的挖掘效率和较低的通信量,适用于对效率要求较高的应用领域。A new distributed association rule mining algorithm of DASM was presented. It adopted the ideas of dynamic itemset counting and sampling, and produced frequent itemsets by meta-learning method. Different sites that applied DASM needn＇t share the same memory. To assure the completeness of the results, the concept of similar degree was introduced. Theory analysis and experiments on the datasets generated using the generator from the IBM Almaden Quest research group show that DASM has better performance and less communication loads. DASM is applicable to those applications where the efficiency could be more important than accuracy results.

关 键 词：抽样 元学习 动态项集计数 相似度 分布式关联规则挖掘 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]