DocumentCode :
441879
Title :
Classification methods based on bipartite graph
Author :
Qi, Heng-Nian ; Wang, Hang-Jun ; Jiang, Zhen-Jie ; Chen, Er-Xue
Author_Institution :
Sch. of Inf. Eng., Zhejiang Forestry Univ., Lin´´an, China
Volume :
4
fYear :
2005
fDate :
18-21 Aug. 2005
Firstpage :
2489
Abstract :
Bipartite graph is a kind of special graph. Generally, it is used for solving matching problems such as assignment problem, fault diagnosis problem and so on. But from the definition, bipartite graph has intrinsic classification implication. The paper overviews some related studies on classifications based on bipartite graph, and then presents a new classification method based on bipartite graph for homogeneous objects. It shows that objects can be classified according to the relations among them. For the relations are decided by the attributes of the objects directly or indirectly. The classification can also reveal the dependence between the relations and the attributes. The correspondent classification algorithm called Big-C is also given in the paper.
Keywords :
graph theory; pattern classification; Big-C classification algorithm; assignment problem; bipartite graph; fault diagnosis problem; intrinsic classification implication; special graph; Bipartite graph; Classification algorithms; Cybernetics; Fault diagnosis; Forestry; Machine learning; Optimal matching; Big-C; Bipartite graph; Classification algorithm; Classifier;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Machine Learning and Cybernetics, 2005. Proceedings of 2005 International Conference on
Conference_Location :
Guangzhou, China
Print_ISBN :
0-7803-9091-1
Type :
conf
DOI :
10.1109/ICMLC.2005.1527362
Filename :
1527362
Link To Document :
بازگشت