American
International
University-
Bangladesh

Students & Supervisors

Abstract

Predicting crop yields accurately is essential to improve agricultural planning, maximize resource allocation, and ensure food security. Predictive models can help farmers and policymakers make data-driven decisions to improve sustain-ability and productivity by using machine learning techniques. To predict crop yields based on agricultural and environmental data, this study uses a variety of machine learning models, such as Ridge Regression, Linear Regression, Random Forest (RF), XGBoost, and Gradient Boosting (GB). The R square (R²), mean squared error (MSE), mean absolute error (MAE), root mean squared error (RMSE), and mean absolute percentage error (MAPE) are used to assess the performance of the model. The most accurate models were Ridge and Linear Regression, which showed a strong linear relationship between features and yield with an R² of 91.23%. Although XGBoost and Random Forest also showed good predictive power, their accuracy was not higher than that of the linear models. Ridge Regression was found to be the most effective model because it strikes a balance between predictability and ease of use, offering insightful information to enhance agricultural planning, resource management, and decision-making in the agricultural industry.

Keywords

Publication Details

• Type of Publication: Conference 
• Conference Name: 2025 IEEE International Conference on Quantum Photonics, Artificial Intelligence, and Networking
• Date of Conference: 31/07/2025 - 31/07/2025
• Venue: Bangladesh Army University of Science and Technology (BAUST), Saidpur, Nilphamari, Rangpur Division, Bangladesh
• Organizer: Bangladesh Army University of Science and Technology (BAUST)