Skip to main content
. 2018 May 29;12:29. doi: 10.3389/fninf.2018.00029

Table 4.

Summary of classification accuracy results achieved with SVM and CNN in the classification experiment leave-one-out participant.

Participant SVM CNN
out Mean ± std Max Min AUC Mean ± std Max Min AUC
P1 44.7 ± 0.6 46.0 43.3 0.44 54.2 ± 1.9 58.2 50.2 0.54
P2 63.1 ± 1.1 65.2 59.7 0.62 65.3 ± 1.8 69.3 58.1 0.74
P3 60.0 ± 0.8 61.9 58.4 0.59 54.6 ± 3.1 60.5 46.9 0.57
P4 57.9 ± 0.9 60.0 55.2 0.57 62.7 ± 1.4 66.2 59.2 0.65
P5 44.8 ± 0.5 46.2 42.9 0.47 50.9 ± 1.7 56.3 47.0 0.50
P6 54.7 ± 0.7 56.2 52.8 0.53 59.2 ± 1.5 63.3 55.0 0.61
P7 53.3 ± 1.2 57.0 50.4 0.54 67.2 ± 1.7 70.8 60.5 0.74
Average 54.1 ± 6.7 65.2 42.9 0.53 59.2 ± 6.0 70.8 46.9 0.62

The lower row shows the grand-average across-all-participants.