| Title of paper:
|
Calibrated circular and elliptical regions for intuitionistic fuzzy assessments: A methodological benchmark of geometry coverage and boundary effects
|
| Author(s):
|
Panagiotis Chountas  0000-0003-3157-8887
|
| School of Computer Science and Engineering, University of Westminster, 115 New Cavendish Street, London W1W 6UW, UK
|
| chountp@westminster.ac.uk
|
| Georgy Urumov 0009-0002-0516-615X
|
| School of Computer Science and Engineering, University of Westminster, 115 New Cavendish Street, London W1W 6UW, UK
|
| G.Urumov1@westminster.ac.uk
|
|
| Published in:
|
Notes on Intuitionistic Fuzzy Sets, Volume 32 (2026), Number 3, pages 258–279
|
| DOI:
|
https://doi.org/10.7546/nifs.2026.32.3.258-279
|
| Download:
|
 PDF (1692 Kb, File info)
|
| Abstract:
|
Circular and elliptic intuitionistic fuzzy sets motivate region-based descriptions of membership and non-membership uncertainty. This methodological benchmark compares circles, axis-aligned ellipses and oriented ellipses on the admissible intuitionistic triangle. It combines established geometric identities and split-conformal calibration with a reproducible evaluation of coverage, admissible area and parameter variability. Six main scenarios and training-size and coverage-level sensitivity analyses comprise 4,200 independent synthetic datasets and 12,600 fits. The axis-aligned ellipse reduces mean admissible area by 53.4 percent in the main axial-anisotropy scenario; an oriented baseline captures corresponding gains for tilted variation. A full-versus-admissible-area diagnostic isolates the effect of clipping: in the boundary scenario, the oriented ellipse has a 28.5 percent mean-area saving relative to the circle before clipping and a 25.5 percent saving after clipping. Coverage uncertainty accompanies the efficiency comparisons. The contribution is an assessment-domain benchmark and methodological synthesis; it introduces neither a new fuzzy-set algebra nor a new conformal method. Real expert behaviour and downstream decision performance remain outside its evidence.
|
| Keywords:
|
Intuitionistic fuzzy assessments, Circular regions, Elliptical regions, Split conformal calibration, Boundary truncation, Synthetic evaluation.
|
| AMS Classification:
|
03E72, 62G15.
|
| References:
|
- Amaral, L. P., Cabezas, L. M. C., Ramos, T. R., & Pereira, G. H. G. A. (2026). Conformal prediction for compositional data. Preprint. Available online at: https://arxiv.org/abs/2511.18141
- Angelopoulos, A. N., & Bates, S. (2022). A gentle introduction to conformal prediction and distribution-free uncertainty quantification. Preprint. Available online at: https://arxiv.org/abs/2107.07511v6
- Atanassov, K. T. (1986). Intuitionistic fuzzy sets. Fuzzy Sets and Systems, 20(1), 87–96.
- Atanassov, K. T. (2020). Circular intuitionistic fuzzy sets. Journal of Intelligent & Fuzzy Systems, 39(5), 5981–5986.
- Atanassov, K. T. (2021). Elliptic intuitionistic fuzzy sets. Comptes rendus de l’Académie bulgare des Sciences, 74(6), 812–819.
- Atanassov, K. T., & Marinov, E. (2021). Four distances for circular intuitionistic fuzzy sets. Mathematics, 9(10), 1121.
- Henderson, I., Mazoyer, A., & Gamboa, F. (2025). Adaptive inference with random ellipsoids through conformal conditional linear expectation. Preprint. Available online at: https://arxiv.org/abs/2409.18508v2
- Johnstone, C., & Cox, B. (2021). Conformal uncertainty sets for robust optimization. Proceedings of Machine Learning Research, 152, 72–90.
- Messoudi, S., Destercke, S., & Rousseau, S. (2022). Ellipsoidal conformal inference for multi-target regression. Proceedings of Machine Learning Research, 179, 294–306.
- Traneva, V., & Tranev, S. (2023). An elliptic intuitionistic fuzzy model for franchisor selection. Annals of Computer Science and Information Systems, 37, 337–343.
- Vovk, V., Gammerman, A., & Shafer, G. (2005). Algorithmic Learning in a Random World. Springer, New York.
- Xu, C., Jiang, H., & Xie, Y. (2024). Conformal prediction for multi-dimensional time series by ellipsoidal sets. Proceedings of Machine Learning Research, 235, 55076–55099.
|
| Citations:
|
The list of publications, citing this article may be empty or incomplete. If you can provide relevant data, please, write on the talk page.
|
|
