Enhancing Agricultural Productivity with Data-Driven Crop Recommendations

Agriculture plays a major role in the sedentary human civilization. Planting the right kind of crop will increase the crop yield. Different machine learning techniques can be used to create recommendation systems. Crop recommendations are given in accordance with the agricultural parameters. A recommendation system can be provided to the farmer to assist him in planting crops with the help of data science. In case of quality prediction, a lot of resources could come to waste. In order to overcome this demerit in existing systems and to improve its accuracy and reliability, Light GBM Machine Learning Algorithm used. Data sets are analyzed and crop recommendations were made based on the content of soil nutrients and environmental factors. Light GBM is a gradient boosting framework that leverages decision trees to enhance model efficiency and minimize memory consumption. The LightGBM algorithm builds a strong, accurate and reliable recommendation system. In this particular model of recommendation system, certain measured parameters like temperature (8-100 degree celsius), humidity (14-100 g.kg^-1), rainfall (10-300mm), pH level (0-10ph) and soil nutrient content (N, P & K) (0-150 Nm^3 , 0-150 Pm^3 , 0-200 Km^3) are considered. The environmental factors humidity and temperature, atmospheric pressure are found using third party API. The output from API can be acquired in the form of JSON and is then displayed in UI. These data help to provide an accurate prediction about the optimal crop that grows in that land. To help the farmers in crop disease management, diseases that can occur for each crop are shown with their possible causes and fertilizers used for their cure are suggested accordingly