I develop and implement statstistical methodology to solve for scientific problems under the advisement of Dr. Vadim Zipunnikov. My major research interest is related to high dimensional and high throughoutput data collected by Wearable Device (accelerometers, actigraphs, heart rate monitor, etc.). Recently, these devices have gained their popularity in both academia and industry; however, accurately analyzing data generated by them remains a big scientific chanllange. Understanding this type of data, their distribution within person and in the population can help understand the association between these new measurements and human health.

My main area of expertise is related to data collection and analysis for objectively measured physical activity, sleep, and circadian rythmicity using actigraphy, acceleromtry etc. I develop novel statistical methogologies, and collaborate on multiple scientific projects as well. Specificly, I am interested in the following directions: 1) feature engineering for raw actigraphy data to summarize noisy and high dimensional physiological signal; 2) dimension reduction on high dimensional actigraphy data; and 3) integrative analysis of features acquired from multiple domains/modalities/devices.

Scientific Interests: wearable devices and their applications in public health (e.g. mental health and aging), physical activity assessment, sleep, circadian rhythmicity.

Statistical Methods: feature engineering for accelerometry signals, dimension reduction, functional data analysis, integration of multiple modalities.

Selected Research Projects

  • Integration of data from domains of physical activity, sleep, and circadian rhythmicity: Implemented JIVE to efficiently deal with multivariate actgraphy features reprsenting multiple domians of physical activity, sleep, and circadian rhtymicity.

  • Processed the NHANES 2003 - 2006 Accelerometery Data: Processed high dimensonal and noisy NHANES 2003 - 2006 accelerometery data to a ready-to-use data package. Provided tutorial analytical pipeline together with codes for future researchers to conduct statistical analysis on NHANES accelerometery data.

  • Novel Statistical Framework to Quantify Activity Fragmentation: Developed a statistical framework to quantify patterns of sedentary/active time accumulation via metrics of fragmentation of bouts and studied their association with mortality in NHANES 2003-2006.

  • Collaborative Research Projects: multiple collaborative research projects including: monitoring hospitalized patients with various activity trackers, sleep fragmetation and its association with cognitive functions, circadian rhythm and its association with mortality etc.

WordCloud Representing My research