Skip to main content

Table 2 Fivefold CV comparison of classifiers

From: dropEst: pipeline for accurate estimation of molecular counts in droplet-based single-cell RNA-seq experiments

Classifier

Sensitivity on CV (%)

Specificity on CV (%)

Stability on class 1 (%)

Stability on class 0 (%)

KDE

90.4 (±0.8)

91.1 (±3.4)

89.8 (±2.9)

97.3 (±1.3)

Random Forest

89.6 (±2.3)

92.7 (±1.6)

87.3 (±7.3)

97.7 (±1.3)

Robust GP

87.7 (±2)

94.8 (±2.2)

85.3 (±0)

99.3 (±0.5)

  1. Mean ± standard deviation values are shown. Here, class 1 is high-quality cells, class 0 is low-quality cells