Intelligent Crop Recommendation Framework Using Hybrid Learning with Macro and Micronutrients.

Abstract

Purpose: Selecting suitable crops according to soil and environmental conditions remains a major challenge in agriculture. Conventional crop selection practices often depend on experience-based decisions and may not effectively consider variations in soil nutrients and climatic conditions, leading to inefficient resource utilization and reduced productivity. It aims to develop a data-driven framework for improving agricultural decision-making.
Study approach:A framework is developed using supervised ensemble machine learning techniques. The proposed approach integrates soil macro-nutrients & micronutrients, soil properties, and climatic parameters for crop suitability prediction. Multiple machine learning methods, including Random Forest, XGBoost, Logistic Regression, and LightGBM, were evaluated. A hybrid stacking ensemble framework was developed by combining multiple base learners with Logistic Regression as a meta-learner. framework performance was assessed using independent testing and stratified 5-fold cross-validation.
Findings: Proposed hybrid stacking framework achieved the highest performance with 96.70% testing accuracy, outperforming individual models. Random Forest, XGBoost, Logistic Regression, LightGBM and Hybrid algo achieved accuracies of 71.70%, 90.15%, 94.35%, 94.55% with highest 96.70% [Hybrid stacking], respectively. The results demonstrate improved prediction capability through ensemble integration.
Limitations: The framework requires further validation using real-world field datasets from diverse geographical regions. Future work can incorporate real-time sensor data for adaptive recommendations.
Practical and social implications: The framework can support farmers and agricultural stakeholders by providing data-driven crop recommendations, promoting efficient resource utilization and sustainable farming practices.
Originality: This study presents a nutrient-aware hybrid ensemble framework integrating macro-micronutrients and environmental parameters for intelligent crop recommendation, providing a foundation for smart agricultural decision-support systems.