Tests of different approaches to EEG sleep analysis, on individual polysomnograpic recordings from the publicly available Sleep-EDF Expanded database. The recordings and annotations should be present in the /data folder. Currently this project contains only classification based on the discrete Hidden Markov model (DHMM) and feature extraction in the frequency domain (different band ratios).
Example command to run the HMM:
$ python src/run.py data/SC4001E0-PSG.edf data/SC4001EC-Hypnogram.edf