
One Ecosystem, Journal Year: 2025, Volume and Issue: 10
Published: May 2, 2025
The conservation of the genetic diversity native maize in Mexico is a priority due to its cultural, agricultural and environmental importance. This study presents development evaluation imanr package, computational tool based on Boosted Ensembles designed automate classification racial complexes maize. Using national database, model was implemented that leverages morphological geographical variables provide precise rapid classifications. methodology included optimisation key parameters through cross-validation, achieving up 90% balanced accuracy Cohen's Kappa coefficient 0.84. These results highlight robustness compared traditional methods, which rely subjective expert judgement require extended times. findings demonstrate package not only surpasses conventional methods terms efficiency, but also offers an accessible for conserving monitoring diversity, aligning with recommendations Global Maize Project (PGMN). Moreover, usability enhanced by developing graphical user interface, allowing non-specialised users fully utilise potential. represents significant advancement science, contributing modernisation identification processes strengthening sustainable management strategies this essential resource. directly addresses need innovative tools monitor preserve suggests promising pathway future applications sector.
Language: Английский