Time Domain Cluster Analysis of Human Activity Using Triaxial Accelerometer Data

Abstract

This paper presents a cluster analysis of raw tri-axial accelerometer data aquired from various human physical activities as well as simulated falls. Clustering was performed using K Means, Gaussian Mixed Model and Fuzzy C-Means clustering. In our analysis we focused on two problems: the first clustering problem was based on activity recognition and differentiation from simulated human falls, while the other problem focused on distinction between single jerk events (e.g. jumping, falling) and continuous activity signals (e.g. running, walking).