Sleep is an important part of life that impacts and is impacted by many processes. Due to the ongoing worldwide increase of sleep disorder diagnosis, it has become even more crucial to study; many disorders can be refected in the sleep-wake cycle. Mice are a popular choice in sleep research due to their similarities to the human sleep-wake cycle. The current golden standard of by-hand visual classifcation of EEGs supplemented with EMG data is timeconsuming, costly, and limited in accuracy by the expert carrying it out. The current research attempts to fnd a new way to classify sleep-wake stages based on only EEG data recorded from mice using an automated algorithm. This research applies epsilon-complexity to the original EEG signal after segmentation by a constant time epoch followed by change-point detection and kmeans classifcation; these methods were previously developed by Piryatinska and Darkhovsky. This process reduces the original EEG from a noisy signal to a pair of denoised diagnostic sequences which are then used in k-means classifcation for separation into two sleep-wake stages: Wake/REM and NREM. Applying this process to a set of publicly available mouse EEGs (n=37) with by-hand sleep scores resulted in an overall accuracy rate of about 86%. This method reduces the time and cost previously associated with this classifcation task. Once the parameters in the method were tuned with this data, the algorithm was then applied to a novel set of EEGs prepared by a lab at UCSF. Future work will focus on further separation of Wake and REM stages. v * wrindex = fnd(sw == 1); * wrsleep = sw(wrindex); * wridx = idx(wrindex); * specn = (sum(wridx == wrsleep))/length(wrsleep); -NREM Sensitvity * nindex = fnd(sw == 2); * nsleep = sw(nindex); * nidx = idx(nindex); * sensn = (sum(nidx == nsleep))/length(nsleep); -NREM Precision * precn = (sum(wridx == wrsleep))/length(nidx);