Table 2 Mean performance indices of the best classification approaches (in %, for 10-cross validation). Se, Sp, Acc – sensitivity, specificity and accuracy, respectively (in %, for 10-cross validation).
Classifier | Heartbeat | Se | Sp | Acc | Classifier | Heartbeat | Se | Sp | Acc |
|---|---|---|---|---|---|---|---|---|---|
k-NN | NOR | 100 | 99.8 | 98.6 | k-NN | NOR | 100 | 100 | 98.3 |
k = 5 | ISM | 97.7 | 99.3 | k = 5 | ISM | 98.2 | 99.2 | ||
AreaD | ISE | 98.2 | 98.5 | AreaRa | ISE | 97.7 | 98.4 | ||
VPB | 95.9 | 99.7 | VPB | 95.6 | 99.8 | ||||
SVM | NOR | 92.7 | 92.6 | 88.8 | SVM | NOR | 93.3 | 92.8 | 86.9 |
RBF | ISM | 80.9 | 95.4 | RBF | ISM | 82.3 | 95.1 | ||
CommonR | ISE | 90 | 96.9 | AreaD | ISE | 85 | 95.4 | ||
VPB | 92.4 | 99.8 | VPB | 87.8 | 98.9 | ||||
NB | NOR | 72.7 | 98.7 | 87.4 | NB | NOR | 97.7 | 82.4 | 83.4 |
Kernel | ISM | 45 | 73 | Kernel | ISM | 41.3 | 95.1 | ||
S50 | ISE | 68.2 | 79.4 | AreaRa | ISE | 38.6 | 97.2 | ||
VPB | 93 | 63.2 | VPB | 94.1 | 82.1 | ||||
DFA | NOR | 100 | 82.2 | 81.7 | DFA | NOR | 91.4 | 95.5 | 77.1 |
Linear | ISM | 49.1 | 97.2 | Linear | ISM | 73.3 | 89.5 | ||
AreaD | ISE | 79.5 | 90.8 | CommonR | ISE | 68.6 | 86.8 | ||
VPB | 75 | 93.7 | VPB | 75.6 | 94.5 |
