Development and Validation of an Ensemble Machine Learning Model for Enhanced Crop Yield Prediction

Abstract

Accurate crop yield prediction is essential towards effective agricultural planning and food security for the growing population. This study aimed to develop and evaluate an ensemble machine learning model for crop yield prediction focusing on improving predictive accuracy and providing actionable insights for agricultural decision-making. The study utilized three machine learning algorithms – Decision Tree, Random Forest, and XGBoost. An ensemble approach using XGBoost was employed to combine the predictions of these algorithms, resulting in an R-squared (R2) value of 0.99, MAE of 608.06 and MSE of 692453.82 showcasing the superior performance of the ensemble model compared to individual algorithms. The ensemble model’s high accuracy demonstrates its potential for improving crop yield predictions. The model was further integrated into a user-friendly android application to assist farmers and agricultural stakeholders in making informed decisions