With the growing amount of IoT farming data, it has become more challenging to select suitable data for IoT farm applications. This study proposes a sophisticated crop recommendation system based on the combination of multiple data sources such as Crop_Recommendation.csv, Soil.csv, and Crop_names.csv that enable to predict one right crop. The system uses geographical coordinates (latitude ϕ and longitude λ) to generate temperature and humidity as environmental variables (expected outcomes of the regression models), which are the basic inputs for the analysis of crop suitability. Using a classification model, with the features such as soil type and nitrogen requirements, the model which was proposed predicts the best crop class. The system tunes the hyperparameters to get the best predictions, and for each scenario it outputs the top five crops that should grow in the best conditions. Also, it computes the Growth Degree Days (GDD) and the nutrient (nitrogen, phosphorus, potassium) needs for all of the suggested crops, allowing the farmer to make an all-encompassing decision. This machine learning and geo-based technique contributes to improved agricultural choices by enabling accurate, data-informed crop predictions that are environment and soil specific, thus is a prospective solution to current related research problems.
Keywords: Crop Recommendation System, Predictive Algorithms, Soil Nutrient Monitoring, Geographic Coordinates, Environmental Modeling, Soil Classification.
How to cite this article: Chintamaneni P, Kodukula S, An efficient Machine Learning based Soil Nutrient Monitoring and Crop Recommendation System. Int J Drug Deliv Technol. 2026;16(4s): 210-227; DOI: 10.25258/ijddt.16.210-227