Figure 1.

ML-Based Classification of “Informative” and “Noninformative” Genes.

(A) Numbers of “noninformative” genes (y axis) in root and shoot tissues at different iterations (x axis) when using the PSOL-based RF classification model.

(B) AUC scores using the PSOL-based RF classification model to classify positive samples (known stress-related genes) and negative samples (“noninformative” genes) in root and shoot tissues at different iteration times.

(C) Influence of the size of the positive samples on the PSOL-based RF classification model.