Generalizability of machine learning models for plant traits using hyperspectral reflectance data: The case of maize
Authors: Xu, R., Ferguson, J. N., Kromdijk, J., Nikoloski, Z.
Category: Plant Biology
Model Organism: Zea mays
▶ AI Summary
The study evaluates hyperspectral reflectance data to predict 25 anatomical, gas exchange, and chlorophyll fluorescence traits in a maize multi-parent recombinant inbred population across three seasons, using various machine‑learning models. It assesses model performance for unseen genotypes, seasons, and combined scenarios, and examines the impact of data aggregation within a rigorous nested cross‑validation framework. The findings highlight trait‑specific limits to model generalizability and provide a robust protocol for applying hyperspectral phenotyping in breeding programs.