Table 3. Actual classification performance for WDBC dataset using KNN classifier.

	Sensitivity	Specificity	GMean	PPV	F1-measure
DWFS	96.86% (±2.68)	89.38% (±4.49)	93% (±1.71)	93.7% (±3.57)	95.2% (±1.89)
MRMR+DWFS	91.38% (±5.71)	85.61% (±5.07)	88.32% (±0.98)	91.35% (±3.92)	91.19% (±2.27)
JMI+DWFS	91.38% (±5.71)	85.61% (±5.07)	88.32% (±0.98)	91.35% (±3.92)	91.19% (±2.27)
MRMR	91.66% (±5.21)	85.61% (±5.07)	88.47% (±0.96)	91.37% (±3.93)	91.36% (±2.06)
JMI	95.31% (±3.26)	88.44% (±3.25)	91.8% (±2.61)	93.41% (±1.35)	94.33% (±1.88)
WEKA	92.2% (±3.95)	86.53% (±2.73)	89.29% (±2.12)	91.8% (±3.03)	91.93% (±2.34)
FST3	95.5% (±3.19)	87.92% (±5.9)	91.57% (±3.07)	92.9% (±4.29)	94.11% (±2.49)
ALL Features	96.16% (±3.59)	88.57% (±3.64)	92.27% (±3.06)	93.43% (±1.97)	94.75% (±2.3)
Correlation-baseline	91.3% (±5.48)	84.05% (±5.3)	87.52% (±3.63)	90.19% (±5)	90.61% (±3.86)