The variable precision fuzzy rough set based on overlap and grouping functions with double weight method to MADM
Zhengqi Shi,Lingqiang Li,Shurui Xie,Jialiang Xie
DOI: https://doi.org/10.1007/s10489-024-05554-3
IF: 5.3
2024-06-14
Applied Intelligence
Abstract:Variable precision fuzzy rough set (VPFRS) is widely utilized for handling various forms of uncertain information due to its fault-tolerant capability. However, a significant number of these rough sets fail to satisfy the inclusion property (lower approximation included in the upper approximation), posing potential risks in applications. Moreover, a common method of constructing the VPFRS is through triangular norms and triangular conorms. But in certain practical applications, the associative law of triangular norms and triangular conorms may not be essential. Overlap functions and grouping functions can effectively avoid this issue. Therefore, to address the limitations of existing models, we introduce the concept of VPFRS based on overlap and grouping functions, and apply it to a real multi-attribute decision-making problem. Firstly, we propose a novel VPFRS leveraging overlap and grouping functions, and demonstrate that it satisfies the generalized inclusion property. This solves the deficiency in VPFRSs not meeting the inclusion property to some extent. Additionally, with the help of the generalized inclusion property, we introduce a new objective method for computing attribute weights. Subsequently, by integrating the merits of the proposed VPFRS model and the PROMETHEE method, we develop a multi-attribute decision-making method with double weight. Finally, the validity of our decision-making method and weight calculation approach is substantiated through comparison and experimental analysis.
computer science, artificial intelligence