Classifier Performance Assay Based on Human Activity Dataset
Contenido principal del artículo
Resumen
Cell Phones are becoming more refined with each new model release. These days phones
ordinarily consolidate numerous sensors for various functions such as GPS tracking,
temperature sensors, accelerometer, gyroscope, and high-determination cameras. Human
Activity Recognition (HAR) has enabled many applications, and because of the data collected
from the HAR, numerous functions can be performed, such as fitness tracking, healthcare,
entertainment and safety, and other functions. There is abundant research that works upon realtime
processing, which causes more power consumption of mobile devices; as we know,
mobile phones are resource-limited devices, and it is a task of activity recognition, so the
process is as accurate as possible. So, the below research uses the smartphone dataset and
various classification models to determine which classifier is more accurate and provides faster
and more accurate results.