Project Crop Yield Prediction at Multiple Spatial Scales with Statistical Machine Learning Program DUIRI - Discovery Undergraduate Interdisciplinary Research Internship Term Summer 2025 Status Accepted Research Area Crop yield modeling, Global sustainability Description With the advancement of agricultural tools and technologies, high-resolution spatial and temporal data are becoming common in modern agriculture. These datasets allow us to learn from the past and adapt to the dynamic challenges of sustainable agricultural production due to changing crop varieties, management practices, weather patterns, and soil conditions in the future. However, there is a gap between the surge in data generation and the expertise to analyze these datasets for practical insights and decision-making. The research objective of this project is to investigate how changing weather patterns and soil conditions impact corn production at both field and regional scales in the United States. The educational objective is to equip undergraduate students with the necessary skills to understand and analyze complex agricultural datasets. The students will be mentored by Dr. Pratishtha Poudel from the Agronomy department in the College of Agriculture and Dr. James Krogmeier from the Electrical and Computer Engineering in the College of Engineering. Dr. Poudel is an agro-ecosystems modeler with a research program at an intersection of crop science, data science, and crop modeling. She works with dynamic process-based, statistical, and machine learning models to understand and predict crop-soil dynamics. Professor Krogmeier studies signal processing algorithms, experimental embedded signal processing systems, and communications to enable novel applications, optimize performance, and lower costs in large-scale sensor networks as applied in transportation and agriculture. In the last decade his research has expanded significantly in the area of agricultural data systems. The ag data work involves colleagues from ECE, Agricultural and Biological Systems Engineering (ABE), Agronomy, Ag Economics, Ag Education, Computer Science, and Food Science. All are working under the umbrella of the Open Ag Technology and Systems (OATS) Center, which was co-founded with Dennis Buckmaster of ABE. Supervisor Pratishtha Poudel Mentor Type of work required The students will be tasked with: a) Compiling and processing datasets, b) Developing predictive machine learning and/or statistical models for corn yield with appropriate feature design and selection for field and regional scales, and c) Translating the model results into agricultural inferences. The datasets to be used in the project are: a) Field data collected from Dr. Poudel's lab during 2024 for field predictions, and b) An open-source global benchmark dataset for regional predictions to which Dr. Poudel actively contributed USA data. Websites Qualifications Student qualifications: Students should be in good academic standing (GPA: >3.5), should have taken at least one programming course, and be familiar with machine learning methods. Students interested in applications of AI in agriculture are encouraged to apply. Credit Hours 0 Weekly Hours 10 (estimated)