Data Analysis for Physical Activity Monitoring

Abstract

Physical activity is essential for humans for maintaining a healthy and comfortable lifestyle. With science and technological advancements, there comes various guidelines for the amount of physical activity a person should perform. Monitoring the physical activity enables us to follow those guidelines and be aware of own activity. Wearable computing is allowing us to track and monitor our own performed physical activities by mostly intrinsic (minimal) interaction. Physical activity monitoring is an emerging research area in wearable computing. Our thesis is about identifying and classifying which activity is being performed. We have used various classifiers and evaluation metrics to validate our classifier models.