Novel Distance Measures of q-Rung Orthopair Fuzzy Sets and Their Applications

Donglai Wang, Yige Yuan, Zhe Liu, Sijia Zhu, Zhifang SunSymmetry 2024 · DOI

TL;DR

This paper introduces two novel distance measures for q-rung orthopair fuzzy sets (q-ROFSs), inspired by the Jensen-Shannon divergence. These measures offer a more nuanced way to quantify differences between fuzzy sets, improving applications in pattern recognition and decision-making by better handling uncertainty.

What Problem Does This Paper Address?

Fuzzy set theory is crucial for modeling uncertainty, but traditional fuzzy sets have limitations in capturing complex uncertainty. Extensions like intuitionistic fuzzy sets (IFSs), Pythagorean fuzzy sets (PFSs), and Fermatean fuzzy sets (FFSs) offer richer representations. Q-rung orthopair fuzzy sets (q-ROFSs) further enhance this by introducing a parameter 'q' that allows for a broader range of uncertainty. However, existing distance measures for q-ROFSs can sometimes produce identical results for distinct sets, limiting their discriminative power in complex scenarios. This paper addresses this by proposing new distance measures that are more sensitive to subtle differences within the q-ROFS framework, specifically leveraging the Jensen-Shannon divergence, which hasn't been extensively used for q-ROFS distance measures.

Methods

The core contribution of this paper is the development of two novel distance measures for q-ROFSs:

  • DJS_2D: A two-dimensional distance measure that considers membership and non-membership degrees.
  • DJS_3D: A three-dimensional distance measure that additionally incorporates hesitancy degrees.

Both measures are inspired by the Jensen-Shannon divergence, a well-established measure of similarity between probability distributions. The authors adapt this concept to the fuzzy set context by treating the membership, non-membership, and hesitancy degrees as components of a generalized "distribution."

The paper also rigorously analyzes the properties of these new measures, proving that they satisfy essential distance measure axioms:

  • Non-degeneracy: The distance is zero if and only if the two fuzzy sets are identical.
  • Symmetry: The distance from set A to set B is the same as the distance from set B to set A.
  • Boundedness: The distance values fall within a specific range (0 to 1).
  • Triangular Inequality: The distance between two sets is less than or equal to the sum of the distances from a third set to each of the first two.

Datasets

The paper does not introduce new datasets. Instead, it uses synthetic examples and numerical experiments to evaluate the proposed distance measures. These examples involve predefined q-ROFSs with varying parameters to test the properties and performance of the new measures.

Key Insights and Ideas

  • Jensen-Shannon Divergence for Fuzzy Sets: Adapting a concept from probability theory (Jensen-Shannon divergence) to fuzzy set theory, particularly for q-ROFSs, provides a powerful new tool for measuring differences.
  • Enhanced Discrimination: The proposed measures offer better discrimination between distinct q-ROFSs compared to some existing methods, which can sometimes yield identical results for different fuzzy sets. This is crucial for applications where subtle distinctions matter.
  • Handling Multi-faceted Uncertainty: The DJS_3D measure's inclusion of hesitancy degrees allows for a more comprehensive representation and measurement of uncertainty, which is a key advantage of q-ROFSs.
  • Parameter 'q' Sensitivity: The paper demonstrates how the parameter 'q' in q-ROFSs influences the distance measures, allowing for flexibility in modeling different levels of uncertainty.

Key Results

  • The proposed distance measures, DJS_2D and DJS_3D, were proven to satisfy the fundamental properties of distance measures (non-degeneracy, symmetry, boundedness, and triangular inequality).
  • Numerical examples demonstrated the effectiveness of the proposed measures in distinguishing between different q-ROFSs, outperforming some existing measures that showed counter-intuitive results.
  • The paper successfully applied the new distance measures to two practical scenarios:
  • Pattern Recognition: The measures were used to classify test samples into designated patterns, showing accurate categorization.
  • Multicriteria Decision-Making (MCDM): The measures were applied to a house selection problem, where they helped rank alternatives and identify the most optimal choice, consistent with other established methods.

Limitations and Future Work

  • Equal Weighting of Attributes: The current distance measures treat all attributes (membership, non-membership, hesitancy) equally. In real-world applications, attributes often have different levels of importance.
  • Future Work Suggestions:
    • Weighted Measures: Integrate weighted information into the distance measures to account for varying attribute importance, enhancing their practicality.
  • Complex Environments: Extend the application of these distance measures to more complex decision-making environments.

Summary updated

BibTeX
@article{wang2024novel,
  author = {Donglai Wang and Yige Yuan and Zhe Liu and Sijia Zhu and Zhifang Sun},
  title = {{Novel Distance Measures of q-Rung Orthopair Fuzzy Sets and Their Applications}},
  journal = {Symmetry},
  year = {2024},
  doi = {10.3390/sym16050574},
}
Figure