Overlap function-based fuzzy β -covering relations and fuzzy β -covering rough set models
Yaoyao Fan,Xiaohong Zhang,Jingqian Wang
DOI: https://doi.org/10.1016/j.ijar.2024.109164
IF: 4.452
2024-03-16
International Journal of Approximate Reasoning
Abstract:As an extension of the fuzzy covering, fuzzy β -covering has garnered significant scholarly concern. However, certain limitations impede its practical application. To address the issue of inaccurate characterization of object relationships caused by the current fuzzy β -neighborhood operator, four new operators were developed, which exhibit both symmetry and reflexivity through the utilization of established fuzzy β -neighborhood operators, overlap functions and grouping functions. Furthermore, we demonstrate that these operators satisfy the fuzzy β -covering relation, and utilize the fuzzy β -covering relations on the basis of overlap functions to propose new fuzzy β -covering rough set model. Additionally, incorporating the attribute significance, an attribute reduction algorithm is designed. Ultimately, we substantiate the rationality and superiority of our proposed algorithm by conducting a sequence of experiments. Meanwhile, we analyze the impacts of varying overlap functions and β values on the algorithm's performance.
computer science, artificial intelligence