Figure 1

Architecture of the convolutional neural network. The input shape (2 second 24-channel EEG) has dimensions 256 (samples) × 24 (channels); the output of the net is dichotomous: 1 (male) or 0 (female). Stochastic optimization was realized using Adamax⁵¹ with learning rate = 0.002, β1 = 0.9, β2 = 0.999, ε = 10⁸ and decay = 0.0. As the loss function, the categorical cross-entropy was used. The total number of parameters was 9,051,902.