Title :
An Efficiency apriori Algorithm: P_Matrix Algorithm
Author :
Bai, Sixue ; Dai, Xinxi
Author_Institution :
Nanchang Univ., Nanchang
Abstract :
Apriori algorithm is the most classical algorithm in association rule mining, but it has two fatal deficiencies: more scans for the databases and generates a great deal of candidate itemsets. To solve these two deficiencies of Apriori algorithm, an efficient and fast algorithm based on Pattern Matrix (named P_Matrix algorithm) is proposed, which solves these two problems: scanning for the databases is cut down to one time and not generating the candidate itemsets but generating the frequent itemsets directly. It greatly reduces the temporal complexity and spatial complexity of the algorithm and highly promotes the efficiency of Aprior algorithm.
Keywords :
computational complexity; data mining; matrix algebra; Apriori algorithm; PMatrix algorithm; association rule mining; pattern matrix; temporal complexity; Algorithm design and analysis; Artificial intelligence; Association rules; Computer science; Data mining; Data privacy; Data warehouses; Itemsets; Spatial databases; Transaction databases;
Conference_Titel :
Data, Privacy, and E-Commerce, 2007. ISDPE 2007. The First International Symposium on
Conference_Location :
Chengdu
Print_ISBN :
978-0-7695-3016-1
DOI :
10.1109/ISDPE.2007.136