CROPUP – A Crop Yield Prediction and Recommendation System with Geographical Data using DNN and XGBoost

Abstract

Agricultural management is significant in a populous country like India. Farmers must have advance knowledge about predicted crop production and crop condition within particular area to make economic and farming decisions. To generate yield, we consider factors like temperature, humidity, pressure, NDVI values, Latitude, Longitude etc. When cultivating a particular crop on a specific type of soil, there are a number of factors to be considered. A crop recommender system considers soil properties such as N, P, and K, as well as other factors like rainfall, humidity, and pH levels, to choose the best crop for the farm. This paper presents a predictive algorithm that would estimate crop yield using deep neural networks with geographical data. A recommendation system was built using machine learning algorithm like Xgboost to recommend the suitable crop. A user interface named CROPUP has been developed to scale up crop productivity and efficiency using the proposed algorithms.