Table 3 Recognition accuracy by n-fold cross validation procedure for different feature extraction techniques for SVM classification for the EDD dataset.
Feature sets | n= 5 | n= 6 | n= 7 | n= 8 | n= 9 | n= 10 |
|---|---|---|---|---|---|---|
PF1 [23] | 50.2 | 50.5 | 50.5 | 50.7 | 50.8 | 50.8 |
PF2 [23] | 49.3 | 49.5 | 49.7 | 49.8 | 49.8 | 49.9 |
PF [24] | 54.7 | 55.0 | 55.2 | 55.4 | 55.5 | 55.6 |
O [21] | 46.4 | 46.6 | 46.6 | 46.7 | 46.7 | 46.9 |
AAC [5] | 40.3 | 40.6 | 40.7 | 40.7 | 40.9 | 40.9 |
AAC+HXPZV [5] | 40.2 | 40.4 | 40.6 | 40.7 | 40.9 | 40.9 |
ACC [29] | 84.9 | 85.2 | 85.4 | 85.6 | 85.8 | 85.9 |
PSSM+PF1 [55] | 74.1 | 74.5 | 74.7 | 75.0 | 75.1 | 75.2 |
PSSM+PF2 [55] | 73.7 | 74.1 | 74.5 | 74.6 | 74.7 | 74.9 |
PSSM+PF [55] | 78.2 | 78.6 | 78.8 | 79.0 | 79.1 | 79.3 |
PSSM+O [55] | 67.6 | 68.0 | 68.1 | 68.3 | 68.3 | 68.5 |
PSSM+AAC [55] | 60.9 | 61.3 | 61.5 | 61.6 | 61.7 | 61.9 |
PSSM+AAC+HXPZV [55] | 66.7 | 67.2 | 67.4 | 67.7 | 67.8 | 67.9 |
Mono-gram [19] | 76.2 | 76.3 | 76.6 | 76.8 | 77.0 | 76.9 |
Bi-gram [19] | 83.6 | 84.0 | 84.1 | 84.3 | 84.3 | 84.5 |
k -AAP (this paper) | 90.1 | 90.2 | 90.4 | 90.5 | 90.6 | 90.6 |