We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Generalized rough set models determined by multiple neighborhoods generated from a similarity relation.
- Authors
Dai, Jianhua; Gao, Shuaichao; Zheng, Guojie
- Abstract
Rough set theory is widely used to deal with uncertainty. Original rough set model is mainly based on equivalence relations. To extend the application scope, classical rough set model based on equivalence relations is generalized to rough set model based on similarity relations. In the present paper, we propose and investigate three new generalized rough set models by introducing new definitions of lower and upper approximations based on multiple neighborhoods generated from a similarity relation. The characteristics of the proposed approximations are investigated. Theoretically, analysis indicates the monotonicity of the corresponding uncertainty measures including accuracy, roughness and approximation accuracy. Experiments indicate that the constructed monotonic measures can be used in attribute reduction.
- Subjects
ROUGH sets; SIMILARITY (Geometry); EQUIVALENCE relations (Set theory); APPROXIMATION theory; MONOTONIC functions
- Publication
Soft Computing - A Fusion of Foundations, Methodologies & Applications, 2018, Vol 22, Issue 7, p2081
- ISSN
1432-7643
- Publication type
Article
- DOI
10.1007/s00500-017-2672-x