Crop Prediction based on Soil and Environmental Features Using Classification Models

Crop prediction plays a pivotal role in agricultural planning and management. This study proposes a comprehensive approach utilizing machine learning models to predict crop types based on soil and environmental features. The objective is to achieve high accuracy, precision, F1-Score, specificity, and recall in classification. We employ Decision Trees (DT), Random Forest (RF), Naive Bayes (NB), k-Nearest Neighbor (kNN), and Support Vector Machines (SVM) as classification algorithms. The dataset comprises soil properties, climatic conditions, and geographical factors relevant to crop growth. Decision Trees and Random Forest exhibit superior performance in handling the complexity of feature interactions and decision boundaries. Naive Bayes provides robust classification based on probabilistic inference, particularly effective with limited training data. The results underscore the efficacy of integrating diverse machine learning techniques for crop prediction. The proposed framework enables stakeholders in agriculture to make informed decisions regarding crop selection and management practices, thereby enhancing productivity and sustainability. The proposed experiments demonstrate remarkable performance across all models with 98% of accuracy. Future work involves the integration of additional data sources and advanced algorithms to further improve prediction accuracy and generalization capabilities.