Logistic Model Tree for Human Activity Recognition Using Smartphone-Based Inertial Sensors

Title: Logistic Model Tree for Human Activity Recognition Using Smartphone-Based Inertial Sensors Author: Heba Nematallah, Sreeraman Rajan, Ana-Maria Cretu Affiliation: Carleton University, Canada Abstract: Human Activity recognition (HAR) systems using sensor data have widespread usage in many real-life applications. In this paper, the Logistic Model Trees (LMT) method was applied for predicting the human motion from smartphone-based inertial sensors. This study aims to demonstrate the capabilities of LMT in obtaining higher prediction rates even under short recognition interval down to only one second. The system is trained and tested on two publically available datasets, namely WISDM and UCI-HAR. The proposed LMT method achieved recognition accuracy 90.86% and 94.02% on WISDM and UCI HAR respectively and between 89.82% - 88.73% during cross-dataset evaluation.