DocumentCode
2144511
Title
Extension of Rough Set Theory Based on Strict Similarity Relation
Author
Chen, Meng ; Xia, Xin
Author_Institution
Dept. of Comput., Nanchang Univ. Nanchang, Nanchang, China
fYear
2010
fDate
14-16 Aug. 2010
Firstpage
666
Lastpage
668
Abstract
From researching tolerance relation and valued tolerance relation, we draw a conclusion that the “similarity” of two objects in incomplete information systems defined by tolerance relation and having missing attribute values is very low. We introduce a new extension of rough set based on strict similarity relation, in which the missing values are considered as “any value”, denoted by *. We regard * as a specific attribute value, two * values are equal, * is not equal to any other attribute values. By strict similarity relation we convert incomplete information systems to complete information systems, the new extension not only overcomes the shortcomings of tolerance relation, but also makes classical rough set theory can be directly used to deal with incomplete information systems.
Keywords
rough set theory; tolerance analysis; complete information systems; incomplete information systems; rough set theory; strict similarity relation; tolerance relation; valued tolerance relation; Approximation methods; Cognition; Data mining; Information systems; Integrated circuits; Rough sets; rough set; strict similarity relation; tolerance relation; valued tolerance relation;
fLanguage
English
Publisher
ieee
Conference_Titel
Granular Computing (GrC), 2010 IEEE International Conference on
Conference_Location
San Jose, CA
Print_ISBN
978-1-4244-7964-1
Type
conf
DOI
10.1109/GrC.2010.20
Filename
5576028
Link To Document