Variety Trial Validation: A Framework to Incorporate on-Farm Data

Cultivar selection may have the greatest impact on profitability of any decision a producer makes. Official cultivar trials provide valuable information but cannot possibly represent all possible production conditions. However, with increasing adoption of yield monitoring in cotton, producers will have additional data available to inform cultivar selection decisions in future years. The goal of this work is to use available datasets and machine learning to develop a framework for modeling the genotype x environment interaction. Ultimately, this work would lead to the development of tools that producers could use to select optimal cultivars for their soil types and expected weather conditions.