Bionote
Giorgio Morales is a PhD candidate (ABD) in computer science at Montana State University and a current member of the Numerical Intelligent Systems Laboratory (NISL). He holds a BS in mechatronic engineering from the National University of Engineering, Peru, and an MS in computer science from Montana State University, USA. His research interests are Symbolic Regression, Explainable Machine Learning, Computer Vision, and Precision Agriculture.
Presentation Abstract
In Precision Agriculture, optimizing fertilizer management is essential for maximizing crop yields and improving agronomic efficiency. Traditional management zone (MZ) approaches focus on within-field variability, but often overlook the impact of fertilizer responsivity on MZ determination. This presentation introduces an MZ clustering method that incorporates fertilizer responsivity, using nitrogen (N) fertilizer-yield response curves. We employ a convolutional neural network to generate N-response curves for each field site, then analyze the shapes of these curves using functional principal component analysis. To refine MZ membership, we apply a genetic algorithm-based counterfactual explanation method that solves a multi-objective optimization problem and identifies the key features influencing cluster assignments. Our results highlight that terrain characteristics, such as slope and topographic aspect, significantly impact MZ membership by affecting fertilizer runoff.
Co-Author: John Sheppard