Ranking approach based on compression transformation and distance factor in Pythagorean fuzzy environment with its application in multi-attribute group decision-making
Deli Zhang,Guijun Wang
DOI: https://doi.org/10.1016/j.eswa.2023.120126
IF: 8.5
2023-04-20
Expert Systems with Applications
Abstract:The Pythagorean fuzzy set is not only a generalization of intuitionistic fuzzy set, but also can deal with decision-making problems with multi-attribute information in a wider area. The main work of this paper is that all Pythagorean fuzzy numbers (PFNs) are mapped to a unit triangle region through compression transformation in the Pythagorean fuzzy environment, the novel distance measure, distance factor and score function are proposed and analyzed, and it is proved that the distance measure satisfies axiomatic conditions and score function meets monotonicity. Main contributions: by comparing the distance between each PFN with the smallest element (0,1) and corresponding hesitation factor, a new score function and its ranking criterion are put forward. In this paper, it is first pointed out that there are contradictions between the existing ranking approaches of PFNs and IFNs through counterexamples, and main reasons for these contradictions are analyzed. Secondly, all PFNs are mapped to the unit triangle of the first quadrant by introducing the compression transformation, and a new distance measure of PFNs is put forward according to the distance meaning in the geometric coordinate system, then it is proved that the proposed distance measure satisfies the axiomatic condition of traditional distance. Finally, the distance factor for PFNs is introduced by comparing the distance between each PFN and the smallest element (0,1), a new score formula and its ranking criterion for PFNs are given, and some properties of the distance factor and score function are obtained. Meanwhile, some advantages of the proposed ranking approach are illustrated by numerical examples, also the approach is applied to multi-attribute group decision-making with multi-attribute information.
computer science, artificial intelligence,engineering, electrical & electronic,operations research & management science