A Time-Series Hybrid Multi-Model Machine Learning Framework for Staple Crops Yield Prediction

Agriculture is pivotal for the economy of a country as it is a major source of food, em-ployment and raw materials. However, challenges such as diseases, soil degradation, and water scarcity persist. Technology adoption can address these issues, improving production and quality. Machine learning enables prediction in agriculture. It opti-mizes irrigation, fertilization, and crop selection, aiding decision-making for food se-curity and crop management. This study proposes multi-model machine learning models for eleven staple (Bananas, Maize, Wheat, Cassava, Rice , Soybeans, Barleys, Potatoes, Beans dry, Peas dry and Cocoa beans ) crop yield prediction. The compara-tive results show that the prediction results of the proposed multi-model algorithm are significantly better than linear model. The error trend seasonality-artificial neural network (ETS-ANN) achieved 80% R2 for Cassava crop yield prediction whereas Ba-nanas achieved lowest R2 (20%).