Crop Prediction Expert System with Ensemble Machine Learning Technique

The activity of growing natural resources for human sustenance and the financial benefit is known as agriculture. After choosing what to grow, farmers will loosen the soil and apply nutrient-rich fertilizers to prepare the field. Sometimes the farmers especially the beginners get confused about what to grow and what amount of nutrients are necessary for the soil. Farmers consistently grow the same crops without attempting new varieties or growing the wrong crop which causes soil acidification. To overcome these problems, in this paper the crop prediction expert system (CPES) is developed that will predict which is the most suitable crop to grow according to some factors like the ratio of nitrogen, phosphorus, and potassium in the soil, the temperature in degrees Celsius, relative humidity in percentage, soil pH, and rainfall in millimeters. The ensemble learning model is used in CPES that combines support vector classifier, random forest classifier, and logistic regression. In this system, the ensemble learning model provides 99% accuracy, precision, recall, and f1 score.