XGBoost Outperforms Traditional Models in Crop Recommendation Accuracy and Reliability

Authors: Harshal Patel, Jitendra Shrivastav, Kamlesh Patidar

Abstract: Crop recommendation plays a vital role in modern agriculture, enabling farmers to make informed decisions that enhance yield and sustainability. With advancements in machine learning, various algorithms have been applied to predict the most suitable crops based on soil and environmental parameters such as nitrogen, phosphorus, potassium, pH, temperature, humidity, and rainfall. This study presents a comparative analysis of several supervised learning models, including Decision Tree, Naïve Bayes, Support Vector Machine (SVM), Logistic Regression, Random Forest (RF), and XGBoost. Performance was evaluated using accuracy, precision, recall, and F1-score. The results demonstrate that while traditional models achieved strong performance, XGBoost outperformed them all, achieving the highest accuracy (99.31%), precision (100%), and reliability across metrics. Its ability to capture complex, non-linear relationships within soil data underscores its effectiveness for precision agriculture. The findings highlight XGBoost as a robust and scalable solution for enhancing crop recommendation systems.