Counterfactual Analysis of Neural Networks Used to Create Fertilizer Management Zones

arXiv:2403.10730v1 Announce Type: new
Abstract: In Precision Agriculture, the utilization of management zones (MZs) that take into account within-field variability facilitates effective fertilizer management. This approach enables the optimization of nitrogen (N) rates to maximize crop yield production and enhance agronomic use efficiency. However, existing works often neglect the consideration of responsivity to fertilizer as a factor influencing MZ determination. In response to this gap, we present a MZ clustering method based on fertilizer responsivity. We build upon the statement …

abstract agriculture analysis arxiv counterfactual cs.lg efficiency however management networks neural networks optimization precision production type