指挥信息系统与技术
指揮信息繫統與技術
지휘신식계통여기술
COMMAND INFORMATION SYSTEM AND TECHNOLOGY
2012年
4期
73-76
,共4页
频繁模式挖掘%FP-growth算法%条件频度基%模式增长
頻繁模式挖掘%FP-growth算法%條件頻度基%模式增長
빈번모식알굴%FP-growth산법%조건빈도기%모식증장
frequent pattern mining%FP-growth algorithm%conditional frequency base%pattern growth
FP-growth算法是当前一种高效的频繁模式挖掘算法,分析了FP-growth算法的性能消耗,引入了条件频度基的数据结构用于存储构建条件FP-tree的所有频度信息,提出了自顶向下与自底向上结合的模式增长方法,该方法避免了原算法在遍历FP-tree时的性能浪费。试验表明,运用了快速模式增长的改进算法的效率比原算法显著提高。
FP-growth算法是噹前一種高效的頻繁模式挖掘算法,分析瞭FP-growth算法的性能消耗,引入瞭條件頻度基的數據結構用于存儲構建條件FP-tree的所有頻度信息,提齣瞭自頂嚮下與自底嚮上結閤的模式增長方法,該方法避免瞭原算法在遍歷FP-tree時的性能浪費。試驗錶明,運用瞭快速模式增長的改進算法的效率比原算法顯著提高。
FP-growth산법시당전일충고효적빈번모식알굴산법,분석료FP-growth산법적성능소모,인입료조건빈도기적수거결구용우존저구건조건FP-tree적소유빈도신식,제출료자정향하여자저향상결합적모식증장방법,해방법피면료원산법재편력FP-tree시적성능낭비。시험표명,운용료쾌속모식증장적개진산법적효솔비원산법현저제고。
The FP-growth algorithm is currently an efficient algorithm for frequent pattern mining.The performance consumption of the FP-growth algorithm is analyzed.A data structure of conditional frequency base is introduced to store all frequency information for building condition FP-tree.The combined up-bottom and bottom-up pattern growth method is proposed to avoid the performance waste when traversing the FP-tree.Experimental results show that the efficiency of the rapid pattern growth is improved than that of the original algorithm.
