Yield Prediction
Yield prediction aims to accurately forecast crop production, informing agricultural decision-making and resource allocation. Current research emphasizes leveraging diverse data sources (remote sensing, weather, soil properties, genetics) with advanced machine learning models, including deep neural networks (DNNs), Bayesian optimization, and graph neural networks (GNNs), to improve prediction accuracy and explainability. These advancements are crucial for optimizing agricultural practices, enhancing food security, and mitigating climate change impacts on crop production, particularly by improving risk assessment and resource management. Furthermore, research is actively exploring methods to address data heterogeneity and imbalance, and to enhance model interpretability.
Papers
Process-constrained batch Bayesian approaches for yield optimization in multi-reactor systems
Markus Grimm, Sébastien Paul, Pierre Chainais
Climate-Driven Doubling of Maize Loss Probability in U.S. Crop Insurance: Spatiotemporal Prediction and Possible Policy Responses
A Samuel Pottinger, Lawson Connor, Brookie Guzder-Williams, Maya Weltman-Fahs, Timothy Bowles