Dana Vrajitoru

This paper evaluates deep learning methods for sleep stage classification from EEG signals, finding that LSTM-based models provide more consistent trustworthiness and calibration than EEGNet.