ENHANCING PREDICTION OF SUITABLE CROP USING NMF AND RANGER OPTIMIZATION TECHNIQUESWITH CNN

Abstract

Accurate prediction of crop is necessary to maximize agricultural output and advance environmentally friendly farming methods. A significant natural resource, soil is essential for flood protection, drought mitigation, food production, and water purification. In order to improve prediction accuracy, this paper proposes a hybrid machine learning (ML) framework that combines the Ranger Optimization Technique with Non-Negative Matrix Factorization (NMF) for dimensionality reduction and feature extraction. High-dimensional agricultural datasets are efficiently broken down into useful components by NMF, and the predictive model is optimized for better performance using Ranger Optimization. Real-world datasets including soil characteristics, meteorological conditions, and historical crop yield data are used to assess the suggested framework. With an accuracy of 91%, experimental results show notable gains in prediction accuracy, computational efficiency, and model robustness when compared to conventional methods. This method offers a potent data-driven precision agricultural solution that facilitates better decision-making, efficient use of resources, and increased food security.