M-health predictive data analysis of daily activities and physiological conditions
Conference Paper
Overview
Research
Additional Document Info
View All
Overview
abstract
The wearable devices technology has evolved dramatically in many respects over the last decade, e.g., in the accuracy
and quality of the data provided, as well as the number of parameters measured. This created new opportunities not only
in the fitness domain, but also in the well-being and general health. Indeed, each day, an ever-increasing number of
people are interested in monitoring their daily habits and activities, as well as their physiological and psychological
condition, aiming to improve their overall quality of life. Such tasks can be realized by the use of predictive tools that can
guide and support individuals during their daily living activities. In this regard, wearable devices produce large volume of
data, which should be properly analyzed in a fast and personalized manner. Furthermore, in order to provide the
individuals with more effective and actionable information, other individuals and health professionals should be involved
in the process. Having this in mind, this paper presents a smart health collaborative approach that considers the extensive
analysis of the individual’s wearable data, as well as a health social network to provide an infrastructure for the
individuals to manage their data and easily interact with others individuals and physicians. The experiments focused on
the data analysis module and considered data from four elderly individuals, wearing a Fitbit Charge HR. A clustering
analysis was performed to identify and characterize patterns in their physiological behaviors and daily activities that were
used for the prediction of the individuals’ daily conditions and activities.
