Comparative Analysis of Machine Learning Algorithms for Wheat Grain Classification using Ensemble Learning Approaches
作者
Shivani Rastogi,Ranjana Sharma
标识
DOI:10.1109/smart66937.2025.11389335
摘要
Determining grain quality, supporting global food security, and establishing fair market pricing, the process of wheat grains classification is very important. In this study, a comparative analysis of multiple machine learning algorithms integrated with ensemble learning techniques was conducted to enhance the accuracy of wheat grain classification. Using a publicly available dataset containing seven geometric features of wheat varieties, five prominent algorithms ANN, SVM, KNN, LR and Random Forest were evaluated. The performance of every model was assessed based on F1-score, recall, accuracy and precision. According to Experimental results, ensemble-based models, Random Forest in particular, performed better than other algorithms with accuracies of classification exceeding 98%. While ANN provided comparable performances with deeper architectures, Kernel-based techniques such as SVM also showed highly competitive results. Simpler models such as KNN and LR achieved high accuracies and served as efficacious baselines. The outcomes of this research highlight the ensemble learning potential for achieving highly robust and reliable classification in agricultural applications. The results of the study significantly contribute to the development of automated wheat grain sorting and quality control systems which can support agriculture and smart farming practices.