DocumentCode :
501172
Title :
Research on Classification Rule Based on Backtracking Ant Colony Algorithm
Author :
Zhipeng, Wang ; Jinguang, Sun ; Da, Yin
Author_Institution :
Sch. of Economic Inf. Eng., Liaoning Tech. Univ., Huludao, China
Volume :
2
fYear :
2009
fDate :
15-17 May 2009
Firstpage :
228
Lastpage :
231
Abstract :
Research on the technology of classification the rules of current data mining. Because of accuracy and validity of the classification rules meet, this paper brings up a new algorithm for classification rules: classification rule based on backtracking ant colony algorithm. The core of the algorithm is: the project ant crawl forward, at the same time the backtracking ants crawl backward. If the backtracking ants find better rule, the project ant go back to the last better attribute node, then research. This paper realizate a test system, and prove that use the algorithm of classification rules can make classification more accurate.
Keywords :
data mining; optimisation; pattern classification; backtracking ant colony algorithm; classification rule algorithm; data mining; Classification algorithms; Computational modeling; Data engineering; Data mining; Inference algorithms; Information analysis; Information technology; Machine learning algorithms; Sun; System testing; backtracking ant volony algorithm; classification rule; data mining;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Information Technology and Applications, 2009. IFITA '09. International Forum on
Conference_Location :
Chengdu
Print_ISBN :
978-0-7695-3600-2
Type :
conf
DOI :
10.1109/IFITA.2009.355
Filename :
5231230
Link To Document :
بازگشت