Random Forest Machine Learning Classifier for Seed Recommendation

Abstract

Crop recommendation system is of due importance to the farmers as well as to the country as it reflects the economic growth of the country . Random forest machine learning classifier proposed and implemented for crop recommendation using visual studio code environment with Python . This model will accept the parameters and based on previous data and current data this will suggest best suitable crop to be cultivated. A data analytics approach with machine learning technology significantly improves the prediction of sustainable crop for the specific year of time, this allows the user to see the exact prediction of crop for the further cultivation. The dataset with sample data of 31 crops each with four attributes such as temperature, humidity, rainfall and soil potential of hydrogen (pH) considered. The random forest (RF) classifier tested to predict the suitable crop and provided 95% accuracy, 95.06 % Precision, 94.90 % recall and F1 score 94.8 % .