Machine Learning (ML) based Human Activity Recognition Model using Smart Sensors in IoT Environment
Navita Navita, Pooja Mittal
Abstract
A major progression in sensor-based technologies has resulted in a fast evolution of the Internet of Things (IoT) applications for developing any real-time monitoring systems. Nowadays, an increasing number of aged people living alone dispersed worldwide, and tracking the status of their health function or activity is necessary. In this paper, an IoT-based human activity monitoring model is proposed to continuously check the activities of aged people via smart sensor-based technologies. In this model, vital data are collected by smart sensors or IoT-based devices, and analysis of that data is done by using different machine learning algorithms for detecting any risk in human activity behavior. After evaluating the proposed model, the SVM has attained the highest accuracy of 98.03% which is highly effective for our purpose. SVM outperformed all machine learning algorithms used for analysis purposes.