Volume 18, No. 4, 2021

A comparative study on various Machine Learning Techniques for Human Activity Recognition and Fall Detection


Neelabh sao , Dr. Sipi Dubey

Abstract

Falling is one of the most serious drawbacks prevailing among the elderly people as well as physically challenged people. It has impressed the researchers in the field of behavioural analysis for healthcare applications. The recent technologies like IoT, sensors placements, wearable devices and so on were contributed to improve the performance of the fall detection systems. To bridge the gap in the mobile technologies, this paper is an extensive survey of fall detection using machine learning algorithms. Initially, we present the scope of the vision based algorithms in brief. The recent techniques are surveyed from the aspects of performance achieved and the limitation. Finally, a comparative study is done to find the issues pertaining in this field. The issues like feature selection, theories of collected data, scope of sensors, privacy and supervised and unsupervised machine learning have to be addressed with the scope of the mobile technologies. Though the researchers have achieved steady progress, this research area still confronts the real-time issues.


Pages: 186-202

Keywords: Fall detection, Feature selection, Machine learning techniques, Mobile Technologies, Wearable devices.

Full Text