An IoT-enabled machine learning framework for planning and optimization of melon cultivation systems

Information from the abstract

The farming industry of melons, which produces over 28 million tons annually, is confronted with major challenges regarding environmental changes and the lack of proper planning tools. In this study, the researchers seek to examine the application of the latest machine learning technology and the Internet of Things in the improvement of the planting of melons. In this model, the researchers employed a data collection methodology that involved the collection of 4,320 data points from September 2024 to November 2025, a period of 14 months. To assess the viability of the proposed model, seven different algorithms were compared: Random Forest, SVM, MLP, LSTM, XGBoost, LightGBM, and CatBoost. In the proposed methodology, enhanced feature engineering has been employed, which has generated 47 features. Rigorous evaluation of the proposed model has been carried out using comprehensive hyperparameter optimization. The proposed model has shown promising results using XGBoost, which has produced R-squared of 0.907, MAE equal to 2.98, RMSE equal to 3.87 for regression and accuracy of 89.1% for classification. Rigorous cross-validation of the proposed model has been carried out using a 5-fold cross-validation method, which has shown good generalization ability mean R-squared equal to 0.901 plus or minus 0.012. SHAP analysis has been carried out, which has shown good interpretability. The proposed model has been implemented in the field, which has shown a yield improvement of 15 to 22% and a reduction in water consumption of 18 to 25%.

Why this record is monitored

This record has an Impact Signal of 74/100 based on recency, source, collaboration, and bibliographic signals. It prioritizes monitoring and is not a judgment of research quality.

Related topics: Smart Agriculture and AI · Greenhouse Technology and Climate Control · Irrigation Practices and Water Management