Skip to main content
. 2021 Feb 5;15:619591. doi: 10.3389/fnins.2021.619591

TABLE 1.

Individual classification performance of LF-CNN for 25 participants.

Subj. ID Number of trials per class Performance on original data
 Result of permutation test
AUCval AUCnaïve AUCensemble
101 104 0.67 ± 0.03 0.60 ± 0.10 0.61 ± 0.07
102 133 0.73 ± 0.04 0.70 ± 0.05 0.72 ± 0.06
103 110 0.67 ± 0.03 0.69 ± 0.06 0.69 ± 0.05
104 123 0.62 ± 0.03 0.56 ± 0.07 0.56 ± 0.07
106 119 0.64 ± 0.04 0.66 ± 0.06 0.63 ± 0.07
107 141 0.66 ± 0.02 0.65 ± 0.04 0.66 ± 0.04
108 100 0.63 ± 0.03 0.55 ± 0.08 0.53 ± 0.05 –
109 95 0.73 ± 0.04 0.78 ± 0.06 0.76 ± 0.07 +
110 150 0.72 ± 0.01 0.63 ± 0.08 0.63 ± 0.04
113 97 0.62 ± 0.07 0.65 ± 0.06 0.64 ± 0.04
114 129 0.71 ± 0.03 0.63 ± 0.08 0.67 ± 0.08
115 101 0.69 ± 0.04 0.65 ± 0.10 0.65 ± 0.10
202 205 0.72 ± 0.03 0.70 ± 0.04 0.73 ± 0.02 +
203 217 0.74 ± 0.03 0.76 ± 0.04 0.75 ± 0.04 +
204 124 0.69 ± 0.02 0.61 ± 0.06 0.65 ± 0.03
213 111 0.69 ± 0.03 0.70 ± 0.07 0.69 ± 0.06
214 70 0.62 ± 0.06 0.55 ± 0.13 0.55 ± 0.14 –
215 177 0.71 ± 0.02 0.70 ± 0.04 0.73 ± 0.06 +
216 81 0.56 ± 0.02 0.54 ± 0.13 0.52 ± 0.15 –
217 157 0.67 ± 0.03 0.65 ± 0.07 0.68 ± 0.09
218 76 0.60 ± 0.05 0.57 ± 0.02 0.54 ± 0.07 –
221 196 0.72 ± 0.03 0.74 ± 0.03 0.76 ± 0.04 +
222 165 0.73 ± 0.04 0.69 ± 0.07 0.69 ± 0.04
223 98 0.69 ± 0.02 0.73 ± 0.07 0.72 ± 0.08
224 137 0.75 ± 0.04 0.71 ± 0.04 0.72 ± 0.04 +
M ± SD 128 ± 38 0.67 ± 0.05 0.66 ± 0.07 0.66 ± 0.07

ROC AUCs computed for the test subsets on time interval –0.2 … 0.5 s using the naïve and ensemble testing procedures are denoted as AUCnaïve and AUCensemble, respectively. AUCval corresponds to mean AUC over cross-validation folds. All AUCs values are presented as M ± SD computed over cross-validation and test folds (for details see section “CNN Testing Algorithms”).

“+” denotes the participants with p < 0.05 on all testing folds (H0 hypothesis was always rejected, i.e., the difference from random performance was found for all folds) for both testing procedures.

“–” denotes the participants with p > 0.05 on all testing folds (H0 hypothesis was always accepted, i.e., the difference from random performance was not found for any fold) for both testing procedures.