Skip to main content

Table 2 Performance of Random Forest on the DG dataset with different sampling methods

From: Velo-Predictor: an ensemble learning pipeline for RNA velocity prediction

  PRE-0 PRE-1 PRE-2 PRE-3 Balanced score
Origin 0.65 0.68 0.57 0.66 0.52
SMOTE 0.8 0.75 0.69 0.67 0.73
NCR 0.67 0.72 0.67 0.75 0.55
ADASYN 0.8 0.76 0.7 0.68 0.73
NearMiss 0.67 0.61 0.46 0.46 0.53
SMOTETomek 0.8 0.77 0.71 0.7 0.75
SMOTEENN 0.83 0.81 0.74 0.4 0.64