Decoding Electro-Priming responses in cotton using an integrated taguchi-machine learning framework for optimizing seed performance
COMPUTERS AND ELECTRONICS IN AGRICULTURE, cilt.256, 2027 (SCI-Expanded, Scopus)
- Yayın Türü: Makale / Tam Makale
- Cilt numarası: 256
- Basım Tarihi: 2027
- Doi Numarası: 10.1016/j.compag.2026.112362
- Dergi Adı: COMPUTERS AND ELECTRONICS IN AGRICULTURE
- Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Aerospace Database, Applied Science & Technology Source, BIOSIS, Compendex, Environment Index, Geobase, INSPEC, Academic Search Ultimate (EBSCO), Engineering Source (EBSCO), Technology Collection (ProQuest)
- Erciyes Üniversitesi Adresli: Evet
Özet
Data-driven optimization of seed treatment technologies is important for improving crop establishment and early seedling performance. Cotton (Gossypium hirsutum L.) is a major industrial crop that requires precise and controllable priming approaches to support uniform germination stress-related biochemical responses. In this study, an electro-priming (EP) framework was evaluated using electric intensity (EI) and exposure time (ET) as controllable input factors. Different germination metrics, seedling vigor, and biochemical traits were collected during the experiment. The data generated were analyzed by ANOVA, followed by Taguchi Design (TD) for model robustness and identifying the most influential factor. Thereafter, the data were modeled using seven regression-based machine learning models using multiple performance metrics to find the best model with high predictive ability. Results of ANOVA revealed that EP modulated oxidative biomarkers without producing statistically significant changes in germination and early seedling vigor. Results were confirmed by the TD, which identified electric intensity (EI) as the most dominant factor, with the highest signal-to-noise (S/N) ratio observed for DPPH and dry weight. ML analysis showed that ensemble-based models, particularly random forest and ExtraTrees, provided the most reliable predictive performance. However, no single algorithm was found optimal across all traits, and predictive reliability was trait-specific. Feature-importance analysis validated the findings of TD by identifying EI as the most dominant predictor across most response variables. The integrated ANOVA-TD-ML framework identified EI as the primary controllable driver of EP responses and provides a basis for AI-assisted optimization of seed treatment conditions to enhance biochemical performance without compromising germination physiology.