Extended Data Fig. 4: Prediction accuracy of annual forecast BCRF models trained using species-specific data. | Nature Plants

Extended Data Fig. 4: Prediction accuracy of annual forecast BCRF models trained using species-specific data.

From: Explainable machine learning models of major crop traits from satellite-monitored continent-wide field trial data

Extended Data Fig. 4

Prediction accuracy of models trained using data to 200 DAS from within species, rather than cross-species models, provide good accuracy for projecting yield in a Canola (N = 608), b Chickpeas (N = 243), c Faba Beans (N = 188), d Lupins (N = 403), e Oats (N = 216), f Field Peas (N = 230), g Lentils (N = 331), h Triticale (N = 266), and i Wheat (N = 2296). Test data are for the latest observed year containing successful trials in each species.

Back to article page