Honors Project On-Campus Access Only
We developed methods to make meaningful clusters of Fitbit sleep data to capture variation in nights’ sleep. Existing dissimilarity measures do not account for the ordering, the timing, and the duration of sleep in successive states, so we propose period-dependent distances between probability distributions adapted for the cyclic nature of sleep. These methods provide unique information about sleep stage data not captured with common sleep summaries and different periods of the night provide unique information about a night’s sleep.
Nguyen, Kieu-Giang, "Good Night?: Clustering Sleep Stage Data" (2020). Mathematics, Statistics, and Computer Science Honors Projects. 50.
© Copyright is owned by author of this document