Impact of sensor data glut on activity recognition in smart environments
Alaa E. Abdel-Hakim, Wael A. Deabes · 2017
In activity recognition applications, rich sensor data may be obviously desirable for performance. However, using much sensor data can lead to undesirable glut. In this paper, we investigate the effect of excessive use of sensor data on the performance of activity recognition. Particularly, we study the effect of using analog temperature sensors data on the accuracy of an HMM-based recognition approach. The performance is comparatively evaluated using confusion matrices before and after including additional temperature sensors.