Enhancing Human Activity Recognition through Machine Learning Models: A Comparative Study

Abstract

This study explores Human Activity Recognition (HAR), a machine learning technique utilized in health monitoring and human-computer interaction. HAR identifies human actions through sensor data from accelerometers and gyroscopes in smartphones and wearables. Key components of this technique include model selection, feature extraction, preprocessing, and data collection to classify activities such as standing, lying, sitting, and walking. Despite its potential, privacy concerns warrant further research for effective deployment. A comprehensive analysis of HAR techniques has been described in this research work.