Ensemble Model for Crop Prediction to Handle Extreme Dry and Wet Climate

The increasing levels of greenhouse gases and climate change, including phenomena such as El Nino, have caused extreme climatic events such as floods and droughts in various parts of the Indian subcontinent, disrupting the cultivable environment for the crops. This unpredictability poses a challenge for the farmers. The proposed alternate crop recommendation system considers temperature, rainfall, soil, and market data to recommend alternate crops that are both sustainable for the climate and fetch higher prices. The proposed system uses stacking regressor with base learners such as Support Vector Regressor (SVR), Linear Regression (LR), and Random Forest Regressor (RFR) for yield prediction. The Gaussian naive Bayesian algorithm is used to calculate the standard precipitation index, which helps in classifying the drought levels. The experimental results show that the proposed system yields 97% accuracy in predicting yields with minimal root mean square error.