Table 5. Actual classification performance for miRNA dataset using KNN classifier.

	Sensitivity	Specificity	GMean	PPV	F1-measure
DWFS	71.79% (±6.3)	98.65% (±0.39)	84.09% (±3.68)	81.07% (±3.64)	76.02% (±4.1)
MRMR+DWFS	69.77% (±6.93)	98.03% (±0.18)	82.62% (±4.11)	74.07% (±2.71)	71.74% (±4.38)
JMI+DWFS	68.49% (±5.81)	97.99% (±0.1)	81.86% (±3.41)	73.25% (±2.26)	70.69% (±3.37)
MRMR	68.27% (±7.79)	98.88% (±0.32)	82.06% (±4.64)	83.34% (±2.88)	74.91% (±5.32)
JMI	57.49% (±6.82)	99.2% (±0.24)	75.41% (±4.47)	85.31% (±4.29)	68.53% (±5.51)
WEKA	67.01% (±16.86)	99.22% (±0.34)	81.03% (±10.46)	86.56% (±8.08)	75.12% (±13.62)
FST3	78.24% (±6.73)	99.41% (±0.16)	88.13% (±3.79)	91.58% (±1.67)	84.25% (±3.96)
ALL Features	59.78% (±7.61)	99.52% (±0.26)	77.01% (±4.77)	91.32% (±3.73)	71.98% (±5.18)
Correlation-baseline	70.12% (±7.14)	98.56% (±0.3)	83.05% (±4.23)	79.7% (±4.37)	74.46% (±4.96)