Vol. 7, Issue 12, Part E (2025)

This research presents a comprehensive ensemble machine learning framework that integrates Support Vector Regression (SVR), k-Nearest Neighbors (kNN), and Random Forest (RF) algorithms to predict paddy crop yields with enhanced accuracy. The study addresses the limitations of individual machine learning models by developing a meta-learning approach using Linear Ridge Regression as the final aggregation layer. Data collection encompassed soil health parameters, historical yield records, and environmental variables from 15 districts within the Chhattisgarh Plains agro-climatic zone. The ensemble framework achieved superior prediction accuracy compared to individual models, demonstrating the potential for data-driven agricultural decision-making in rice cultivation systems.