Table 1.

Recognition accuracy by n-fold cross validation procedure for different feature extraction techniques for SVM classification for the DD-dataset.

Feature sets	n = 5	n = 6	n = 7	n = 8	n = 9	n = 10
PF1 [23]	48.6	49.1	49.5	50.1	50.5	50.6
PF2 [23]	46.3	47.0	47.5	47.7	47.9	48.2
PF [24]	51.2	52.2	52.6	52.9	53.4	53.4
O [21]	49.7	50.4	50.8	50.8	51.1	51.0
AAC [5]	43.6	43.9	44.2	44.8	44.6	45.1
AAC+HXPZV [5]	45.1	46.2	46.5	46.8	46.9	47.2
ACC [29]	65.7	66.6	66.8	67.5	67.7	68.0
PSSM+PF1 [55]	62.5	63.2	63.7	64.2	64.5	64.6
PSSM+PF2 [55]	62.7	63.3	64.1	64.2	64.6	64.7
PSSM+PF [55]	65.5	66.2	66.5	66.9	67.1	67.5
PSSM+O [55]	62.5	62.1	62.5	62.9	63.4	63.5
PSSM+AAC [55]	57.5	58.1	58.4	58.7	59.1	59.2
PSSM+AAC+HXPZV [55]	55.9	56.9	57.1	57.7	58.0	58.2
Mono-gram [19]	67.7	68.4	68.6	69.1	69.4	69.6
Bi-gram [19]	72.6	73.1	73.7	73.7	74.1	74.1
k-AAP (this paper)	74.3	75.2	75.2	75.7	76.1	76.1