An Embedded Sensing Methodology for the Classification of Activity Rate

A lack of physical activity can lead to serious injuries, especially for the elderly population. In order to guarantee a high quality of life, the evaluation and monitoring of activity rate require suitable low-cost and wearable solutions. In this paper an embedded sensing solution is proposed, to implement a methodology for the real-time identification and classification of physical activity in three classes of intensity: sedentary, moderate and intensive. To such aim, Rule-Based and Machine Learning algorithms have been assessed on an emulated dataset collected with a supporting structure, by using dedicated metrics. Obtained results demonstrated good performances of the Rule-Based approach, leading to an Accuracy and <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\boldsymbol{F}_{\mathbf{1}}$</tex> -score of 98.77 % and 98.25 %, respectively.