Deep Learning for Classifying Physical Activities from Accelerometer Data

Physical inactivity increases the risk of many adverse health conditions, including world’s major non-communicable diseases, such as coronary heart disease, type 2 diabetes, and breast colon cancers, shortening life expectancy. There are minimal medical care personal trainers’ methods to monitor a patient’s actual physical activity types. To improve monitoring, we propose an artificial-intelligence-based approach classify movement patterns. In more detail, employ two deep learning (DL) methods, namely feed-forward neural network (DNN) recurrent (RNN) for this purpose. We evaluate models on datasets collected from several volunteers who carried tri-axial accelerometer sensors. The first dataset is UCI machine repository, which contains 14 different activities-of-daily-life (ADL) 16 single wrist-worn accelerometer. second includes ten other ADLs gathered eight placed sensors their hips. Our experiment results show that RNN model provides accurate performance compared state-of-the-art in classifying fundamental patterns with overall accuracy 84.89% F1-score 82.56%. indicate our method doctors trainers promising way track understand activities precisely better treatment.