Todim-Based Complex Spherical Fuzzy VIKOR Method for Evaluating the Optimal Band Interval for Classification of Cashmere and Wool Fibers in near-Infrared Spectroscopy
To improve the classification of high-value fibers such as cashmere and wool, which exhibit highly similar structural characteristics, this study proposes a novel multi-attribute group decision-making (MAGDM) framework based on complex spherical fuzzy sets. The proposed method integrates the Complex...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Taylor & Francis Group
2025-12-01
|
| Series: | Journal of Natural Fibers |
| Subjects: | |
| Online Access: | https://www.tandfonline.com/doi/10.1080/15440478.2025.2519667 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | To improve the classification of high-value fibers such as cashmere and wool, which exhibit highly similar structural characteristics, this study proposes a novel multi-attribute group decision-making (MAGDM) framework based on complex spherical fuzzy sets. The proposed method integrates the Complex Spherical Fuzzy Power Maclaurin Symmetric Mean (CSFPMSM) operator with a hybrid TODIM-VIKOR model to address uncertainty, attribute correlation, and extreme values in near-infrared (NIR) spectral data. The CRITIC method is employed to determine objective attribute weights without prior knowledge. Experimental evaluations were conducted using NIR spectroscopy data across four characteristic band intervals (920–930 nm, 1010–1020 nm, 1320–1330 nm, and 1410–1420 nm) for both sliced and whole cashmere and wool fibers. The proposed CSF-TODIM-VIKOR model achieved an F1-score of 0.893, significantly outperforming baseline methods such as logistic regression (0.832), decision tree (0.818), and KNN (0.797), with statistical significance (p < 0.001). Sensitivity and comparative analyses further demonstrated the model’s robustness under varying parameter conditions and its superiority over existing MAGDM approaches. The results confirm that the proposed method effectively reduces the influence of spectral redundancy, improves classification reliability, and enhances decision interpretability under complex fuzzy environments, offering a promising tool for real-time fiber identification in textile manufacturing. |
|---|---|
| ISSN: | 1544-0478 1544-046X |