A Cloud Based Machine Intelligent Human Activity Recognition System Using Internet of Things to Support Elderly Healthcare

Abstract

Purpose: Human activity recognition is now a major concern in elderly healthcare perspective. Regular monitoring of daily activities is strongly needed for the elderly or old age persons at home. Therefore, the Internet of Things (IoTs) can be a solution to this problem. Design/Methodology/Approach: In this paper, a cloud-based machine intelligent human activity recognition (HAR) system using IoT is proposed to recognize the regular activity of old person at home. In this system, the IoT device or wearable device connected to the body is embedded with activity recognition sensors those sense the physical activity and send the readings to the device. The device then sends the readings to the cloud using the Internet for classifying the actual activity of the person. The cloud is installed with a machine intelligent model which accurately classifies the activities. For the selection of this model, in this work we considered many standards supervised machine intelligence models. Findings/Result: Simulation is done using Orange 3.26 python-based tool by considering Kaggle activity recognition data. Results state that NN shows better performance than other models in classifying the activities of the elderly person. Originality/Value: A new cloud-based machine intelligent HAR system for smart home using IoTs is proposed to monitor the regular activity of the old person. Paper Type: Methodology Paper.