Digital sensors in smartphones and wearables measure huge quantities of data, such as heart rate (HR) and accelerometry time series.
However, simple metrics such as resting HR and total step count do not describe structure, complexity, or how information varies with time scale or transfers between different signals.
I did my PhD in the Clifford Lab. We used machine learning, statistics, and information theory to show classification of illness using features from HR and activity time series is improved by considering information over several time scales and between signals.
We developed, patented, and licensed an algorithm to monitor PTSD via cardiovascular stress.
My papers are available via Google Scholar.