摘要
提出了一种挖掘最大频繁模式的有效算法SFP-MFP,给出了最大频繁模式树MFP-TREE的定义,并使用SFP-TREE结构存储挖掘结果,采用了有效的子集检查方法,极大地降低了算法的时空开销,提高了挖掘效率.理论分析和实验表明,该算法的执行效率较其他同类算法有明显改进.
An efficient algorithm for mining maximal frequent patterns （SFP-MFP）, based on sorted FP-TREE, is proposed. Maximal frequent patterns tree （MFP-TREE） is defined, A structure of SFP-TREE, which replaces MFP- TREE,was used to store all maximal frequent item sets, and some subset checking approaches were adopted to im- prove it. Therefore, the proposed algorithm greatly cut down the cost of space and memory, and improved the mining efficiency. Theoretical analysis and experimental results show that the executing performance of the new algorithm proposed is better than other similar algorithms.
出处
《广东工业大学学报》
CAS
2009年第2期64-68,共5页
Journal of Guangdong University of Technology
基金
国家科技支撑计划子课题(2006BAI08B01-03)
