Model Selection and Evaluation: Ensuring Robust and Accurate Prediction of Crop Yields in Agriculture

The focus of this chapter is to introduce some of the key concepts in model selection and evaluation necessary for building Machine Learning (ML) models for improving prediction accuracy and interpretability. We first overview the relevant stages in building and evaluating a ML model, then we present the necessary ML preliminaries, followed by the key techniques and concepts in model selection and regularization. Finally, we explain how to apply these concepts of model selection and evaluation for building ML models for predicting crop yield using weather, soil and agricultural data.