DocumentCode :
424121
Title :
Frequent closed itemsets lattice used in data mining
Author :
Cheng, Zhi-Hua ; Jia, Lei ; Ren-Qing Pei
Author_Institution :
Sch. of Mechatronics & Autom., Shanghai Univ., China
Volume :
3
fYear :
2004
fDate :
26-29 Aug. 2004
Firstpage :
1745
Abstract :
Association rules, classification rules and clustering rules are three main classes of useful rules in the fields of data mining. We often use different algorithms to mine them. In the past few years, the technology, frequent closed itemsets mining, is introduced. It generates a small set of rules compared with the traditional frequent itemset mining without information loss. In this paper, based on the theory of Galois connection, we introduce a new unified structure to mine the three different kinds of rules. The structure is called frequent closed itemsets lattice. We only need to add additional simple function to the algorithm that builds the structure to carry out the process of mining. We find the new structure as very useful and promising.
Keywords :
Galois fields; data mining; learning (artificial intelligence); pattern classification; pattern clustering; Galois connection theory; association rules; classification rules; clustering rules; data mining; frequent closed itemsets lattice structure; frequent closed itemsets mining technology; rule generation; Association rules; Clustering algorithms; Data mining; Electronic mail; Itemsets; Lattices; Mechatronics; Roentgenium; Tellurium; Transaction databases;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Machine Learning and Cybernetics, 2004. Proceedings of 2004 International Conference on
Print_ISBN :
0-7803-8403-2
Type :
conf
DOI :
10.1109/ICMLC.2004.1382057
Filename :
1382057
Link To Document :
بازگشت