Publisher
University of Tennessee at Chattanooga
Place of Publication
Chattanooga (Tenn.)
Abstract
Smartphones play a crucial role in sensor-based human activity recognition (HAR) systems, allowing tracking and measuring human movement. These devices are equipped with sensors, including accelerometers, gyroscopes, and magnetometers, which collaborate to capture different aspects of movement and orientation. Utilizing the data gathered from these sensors, a smartphone's IMU sensor can offer valuable insights into an individual's motion, orientation, and spatial position. Consequently, this information becomes instrumental in recognizing and categorizing activities such as walking, running, cycling, or sitting. This work used various ML models, Support Vector Machines (SVM), K-Nearest Neighbors (KNN), and Random Forest (RF), to detect human activities such as walking upstairs, walking downstairs, standing, and sitting. KNN and SVM achieved the highest accuracy at 95%.
Document Type
posters
Language
English
Rights
http://rightsstatements.org/vocab/InC/1.0/
License
http://creativecommons.org/licenses/by/4.0/
Recommended Citation
Ghaffar Nia, Nafiseh and Qin, Hong, "Towards Implementing Machine Learning Models in Human Activity Recognition". ReSEARCH Dialogues Conference proceedings. https://scholar.utc.edu/research-dialogues/2023/proceedings/11.
Towards Implementing Machine Learning Models in Human Activity Recognition
Smartphones play a crucial role in sensor-based human activity recognition (HAR) systems, allowing tracking and measuring human movement. These devices are equipped with sensors, including accelerometers, gyroscopes, and magnetometers, which collaborate to capture different aspects of movement and orientation. Utilizing the data gathered from these sensors, a smartphone's IMU sensor can offer valuable insights into an individual's motion, orientation, and spatial position. Consequently, this information becomes instrumental in recognizing and categorizing activities such as walking, running, cycling, or sitting. This work used various ML models, Support Vector Machines (SVM), K-Nearest Neighbors (KNN), and Random Forest (RF), to detect human activities such as walking upstairs, walking downstairs, standing, and sitting. KNN and SVM achieved the highest accuracy at 95%.