Fig. 3From: Chronic Kidney Disease stratification using office visit records: Handling data imbalance via hierarchical meta-classificationPlots depicting the performance per-class, measured in terms of Sensitivity (left) and F-measure (right), of the random forests baseline classifier (RF-Baseline, shown in gray), and of over-sampling using SMOTE (RF-SMOTE shown in horizontal stripes) compared with the random forests based hierarchical meta-classifier we have developed (RF-Hier-MC, shown in black). The X-axes in both plots denote CKD stages; the Y-axes indicate the sensitivity (left) and the F-measure (right) per stageBack to article page