Farm Fusion: Smart Crop Recommendation System for Sustainable Farming Using Machine Learning Techniques

In the recent past, data-centric solutions have been developed to address crop choice issues, low yields and soil management. The proposed work in this paper is a web-based application, referred to as Farm Fusion designed to help the farmers for the selection of the best crops to be cultivated according to specific site characteristic. This system utilizes the following machine learning algorithms like Support Vector Machine (SVM), K-Nearest Neighbor (KNN), Random Forest, Gaussian Naive Bayes (GNB) and gradient boosting models LightGBM, XGBoost, and CatBoost. Before suggesting the appropriate crops, the system uses an Ensemble model that integrates these algorithms. The input soil nutrient parameters included in this system are phosphorus, potassium, nitrogen while the input environmental parameters include pH, temperature, rainfall and humidity. It is a web application that uses the flask framework which is integrated with pre-trained model to recommend right crop for the farm field. This system will assist in the process of decision making process, provide sustainability and improve crop yield.