A0326
Title: Two-dimensional functional mixed-effect model for repeatedly measured wearable device data
Authors: Xinyue Li - City University of Hong Kong (Hong Kong) [presenting]
Abstract: With the rapid development of wearable device technologies, advanced accelerometers can record minute-by-minute physical activity for consecutive days. As the daily routine varies throughout the week, accelerometer data can be considered as repeatedly-measured functional observations, which smoothly vary over longitudinal visits with covariate-dependent mean and covariance functions. An innovative two-dimensional functional mixed-effect model (2DFMM) is proposed to characterize the "spatial" (longitudinal) and "temporal" (functional) structures, incorporating two-dimensional fixed effects for covariates of interest. A fast three-stage estimation procedure is also developed to provide accurate fixed-effect inference for model interpretability, and computational efficiency is improved when encountering large datasets. Extensive simulation studies are conducted to demonstrate the effectiveness of the proposed method in comparison with existing approaches. The proposed 2DFMM was further applied to Shanghai school adolescent accelerometer data, demonstrating the effectiveness and computational efficiency of 2DFMM in providing interpretable intraday and interday dynamic associations between physical activity and mental health assessments among Shanghai school adolescents, which further shed light on possible intervention strategies targeting daily physical activity patterns to improve school adolescent mental health.
