Effectiveness of Machine Learning on Human Activity Recognition Using Accelerometer and Gyroscope Sensors: A Survey
Munid Alanazi, Raghdah Saem Aldahr, Mohammad Ilyas
Proceedings of the 26th World Multi-Conference on Systemics, Cybernetics and Informatics: WMSCI 2022, Vol. II, pp. 1-5 (2022); https://doi.org/10.54808/WMSCI2022.02.1
|
The 26th World Multi-Conference on Systemics, Cybernetics and Informatics: WMSCI 2022
Virtual Conference July 12 - 15, 2022 Proceedings of WMSCI 2022 ISSN: 2771-0947 (Print) ISBN (Volume II): 978-1-950492-65-7 (Print) |
|
Abstract
HAR is defined as using machine learning to classify certain human activities in specific time periods by learning from inertial sensor time series data [2]. Over the past few years, the growth in the computational field has been robust enough to transfer a world to a more intelligent place. Currently, the electronic parts turn out to be available in many shapes and sizes. For example, accelerometer and gyroscope sensors could be manufactured into a single piece that can be operated in wired or wireless settings (Bluetooth). Also, this piece could be used in smartphones because the smartphone cannot operate most of its features without these two sensors with respect to the other sensor (magnetometer and inclinometer). The amount of data captured from these sensors in a time series format is in billions of bytes. Human activity recognition is one of the important applications that could be implemented by using these sensors blended with certain machine learning algorithms. Such applications have become important research area because these serve athletic, healthcare, and personal use. This paper summarizes the important aspects of machine learning, human activity recognition, and reviews existing scientific literature in the field of human activity recognition.
|
||