Title: Nonparametric confidence band for activity profiles based on wearable device data
Authors: Hsin-wen Chang - Academia Sinica (Taiwan) [presenting]
Ian McKeague - Columbia University (United States)
Abstract: The motivation comes from applications to health care monitoring in which there is a need to analyze activity profiles constructed from wearable device data. We introduce a nonparametric likelihood ratio approach that makes efficient use of the activity profiles to provide a confidence band for their means. The procedure is calibrated using bootstrap resampling. A simulation study shows that the proposed procedure outperforms competing Wald-type functional data approaches. We illustrate the proposed methods using wearable device data from an NHANES study.