A Novel Patient - Centric Intelli Care System

Abstract

In recent years, mobile health (mHealth) has emerged as a powerful paradigm for real-time monitoring and analysis of physical activities using wearable sensors and smart devices. Human Activity Recognition (HAR) plays a pivotal role in this domain by enabling automated classification and prediction of user activities, contributing to improved healthcare services. This paper explores the development of an intelligent HAR system using the MHEALTH dataset, employing both machine learning and deep learning techniques. Specifically, a Random Forest (RF) classifier and a Long Short-Term Memory (LSTM) neural network were implemented and evaluated for their classification capabilities. Experimental results demonstrate that the RF model achieved superior performance with an accuracy of 97.13%, while the LSTM model attained an accuracy of 91.44%. The study highlights the significance of combining temporal and spatial features for efficient activity recognition and provides insights into model selection for real-world deployment in mobile health applications.