“…All experimental result tables report these 4 metrics as well as areas under the ROC/PR curves. 95% standard-errorbased confidence intervals of performance metrics, which are used in all figures/tables, were constructed by drawing 10 randomized train/validation/test splits (proportion 40:20:40%) with respect to complete recording Energy in frequency bands [0,1], [1,2], [2,3], [3,6] Training, validation and test sets contained 49, 25, and 49 segments, respectively. Splits were stratified, such that the label prevalence difference between training, validation, test sets does not exceed 0.015 in each split.…”