New Pythagorean fuzzy-based distance operators and their applications in pattern classification and disease diagnostic analysis

• Published in: Neural computing & applications Vol. 35; no. 14; pp. 10083 - 10095
• Main Authors: Ejegwa, Paul Augustine, Feng, Yuming, Tang, Shuyu, Agbetayo, Johnson Mobolaji, Dai, Xiangguang
• Format: Journal Article
• Language: English
• Published: London Springer London 01.05.2023; Springer Nature B.V
• Subjects: Artificial Intelligence; Computational Biology/Bioinformatics; Computational Science and Engineering; Computer Science; Data Mining and Knowledge Discovery; Decision making; Deep learning; Fuzzy logic; Fuzzy sets; Image Processing and Computer Vision; Machine learning; Modeling and Utilization; Operators; Parameters; Pattern classification; Performance indices; Probability and Statistics in Computer Science; Representation; S.I.: Interpretation of Deep Learning; Special Issue on Interpretation of Deep Learning: Prediction; Intuitionistic fuzzy set; Pythagorean fuzzy distance; Pythagorean fuzzy set; Disease diagnosis; Pattern classification
• ISSN: 0941-0643; 1433-3058
• DOI: 10.1007/s00521-022-07679-3

Pythagorean fuzzy set is a notion that takes a broader view at intuitionistic fuzzy set with a higher prospect of applications since it has a wider scope compared to IFS. On the other hand, distance measure is an efficient information measure for decision-making via deep learning approach, and thus, this paper proposes a new tri-parametric distance and its weighted version under Pythagorean fuzzy environment with improved performance indexes compared to the hitherto tri-parametric distance measures in the literature. The proposed Pythagorean fuzzy-based distance operators are new approaches because they take into account the three conventional number of parameters of PFSs against the existing practice, and as well incorporate the whole parameters to avoid error due to exclusion as witnessed in other distance operators. Some theorems are presented to validate the new Pythagorean fuzzy distance techniques with regards to its alignment with distance operator’s properties. We demonstrate the applications of the proposed Pythagorean fuzzy distance and its weighted version in cases involving pattern classification and disease diagnosis via deep learning approach where patterns, diseases and patients are presented as Pythagorean fuzzy values. Finally, some comparative analyses of the new tri-parametric Pythagorean fuzzy distance and its weighted version alongside some similar existing distances are presented in terms of the applications to showcase the superiority of the present Pythagorean fuzzy distance techniques.