Table 1. Classification profiles of different network structures.

True positive (TP), false positive (FP), true negative (TN) and false negative (FN) numbers after different inputs or network structures were applied for deep neural networks. Raw data: classification results of a classifier analyzing raw electroencephalogram inputs. Periodogram 1st: classification results of the first neural network analyzing periodogram. Periodogram dual: classification results of sequential application of the first and the second networks analyzing periodogram. Cut-off: cut-off threshold for classifying the input segment to a seizure. The two cut-off values shown in “Periodogram dual” indicate the cut-off for the first and the second network, respectively.