Publication

Abstract : The backbone of our country is Agriculture. For agriculture, soil is a substantial component. Each crop grows better in each soil. Now a day’s intercrop culture is also blooming vastly. Need of encouragement is essential for our farmers to choose correct crop for the soil pattern and suitable intercrop for the main crop. Different soil types can have dissimilar landscapes and dissimilar varieties of yields grow on diverse categories of soils. The growth of crops is affected by the chemical features of soil. It’s essential to identify the geographies and features of numerous soil kinds to recognize which yields produce well in some soil categories and region. Prediction will supportive in choosing the main crop and suitable intercrop for the region. In recent years, Machine learning has advanced a lot. It’s an emergent and thought-provoking explore arena in agronomic data investigation and harvest prediction. In this proposed methodology, classification of soil series according to region then as per the classification the model will predict suitable crop and its intercrop for that particular region. Numerous machine learning classification algorithms are used. For crop prediction and intercrop prediction can be done by using weighted k-Nearest Neighbor, Naive Bayes, Gaussian Support Vector Machine (SVM), XGBoost Algorithms. Also combined the algorithms by Ensemble Learning approach for better prediction. Finally Investigational outcomes show that the proposed XgBoost, SVM and Bagging Ensemble approach achieves improved results than various present methods.