Crop Yield Prediction Using Ensemble Machine Learning Techniques

In order to make strategic decisions about import-export policies and triple farmers’ income, early and precise crop output evaluation is crucial in the agriculture sector. Crop yield predictions are made using machine learning (ML) algorithms, however the agriculture industry faces difficulties with them. Using three tree-based classifiers—Decision tree (DT), Extra tree (ET), and Categorical Boosting (CatBoost)—this study investigated the application of machine learning approaches to forecast agricultural yields. In order to assess the model’s performance, assessment measures such R-squared, Mean Squared Error (MSE), Mean Absolute Error (MAE), and Accuracy were used. A thorough depiction of pertinent agricultural parameters, including weather patterns, soil properties, and past crop yields, may be found within the Kaggle database. According to the results, the ET classifier performed better than the DT and CatBoost classifiers, with the greatest accuracy of 99.15%. These results highlight the effectiveness concerning machine learning techniques in precisely predicting the output of crops, providing insightful information for agricultural decision-making and resource optimization.