Fig. 6: Bayesian sleep-stage classification.
From: Sleep as a random walk: a super-statistical analysis of EEG data across sleep stages
a, b Two examples of automated Bayesian sleep stage classification. In each case, the upper hypnogram shows, for each 30 s epoch, the posterior probabilities of the sleep stages, with larger color density corresponding to larger probability. The middle hypnogram shows only the predicted sleep stage with maximum posterior probability. The lower hypnogram is the ground truth, provided by the specialist human rater. The accuracies are defined as the ratio of correct sleep stage predictions.
