Table 4.

Performance comparison of different methods on Human dataset.

Model	ACC (%)	SN (%)	MCC (%)
LDA + RF [37]	96.4	94.2	92.8
LDA + RoF	95.7	97.6	91.8
LDA + SVM	90.7	89.7	81.3
AC + RF	95.5	94.0	91.4
AC + RoF	95.1	93.3	91.0
AC + SVM	89.3	94.0	79.2
Our method	96.21	94.77	92.70

LDA: Linear discriminant analysis; RoF: Rotation forest; RF: Random forest.