Machine Learning based Crop Suitability Prediction and Fertiliser Recommendation System
B Gayathiri, P. Brindha, I. Karthika, E. Saranya, G Rajeshkumar, P. Rajesh Kanna
Abstract
As the world population grows demand for sustainable agriculture increases. It causes major impacts like rising supply chain costs, not being able to maintain the safety and quality of your products, insufficient communication between parties failure to track and control inventory in warehouses and stores on the food supply chain. Yield forecast which is important for agriculture stakeholders can be obtained by using machine learning models and data from multiple sources. In the previous years, prediction of yield was done manually with the experience of farmers and it was not so accurate. The yield prediction being a major can be solved by the advanced technologies and with the available data. This study offers a system that advises crops appropriate for the region and fertilizers that are suited for crops based on soil measurements to farmers, all based on the yield history of previous years. This agricultural yield forecast and fertilizer recommendation uses machine learning methods. The Random Forest method is used to propose crops since it produces better results than other algorithms, while the K-means clustering technique is used to suggest fertilizers depending on the NPK level of the soil.