Machine Learning-based Approaches for Crop Recommendations and Prediction

The Crop recommendations are crucial nowadays for maximizing agricultural yield. This paper presents extensive research on the application of machine learning techniques for crop recommendation. In this study, machine learning approaches have been proposed to recommend the crop for given soil properties and atmosphere by manipulating the data available from diverse sources. The proposed method includes the study of six machine learning techniques: linear regression, support vector machines, random forests, naive Bayes,
$\mathrm{K}$
-nearest neighbors, and light gradient boosting machines. The framework used to carry out this study includes the data-processing, feature selection, and model training to produce accurate and robust crop recommendation models. The performance of various models is verified using precision, recall, F1-score, and accuracy. The results demonstrate that the Naïve Bayes and random forest outperforms other models with
$99{{\% }}$
accuracy.