Classification of human movements using machine learning: The example of smartphone sensors

Devices that emerged with the developments in wearable technology have quickly become a part of everyday life. Especially the sensors they have increase the usefulness of these devices. Human Activity Recognition (HAR) classifies a person's activity using sensitive sensors that are affected by human movements.Since the Human Activity Recognition feature is on smartphones, it increases both the users' experience and the capabilities (sensors)of the smartphones, and users carry their smartphones with them for this reason. These facts make IAT more important and popular.Human Activity Recognition (HAR) has multiple applications due to the global use of capture devices such as smartphones and video cameras and their ability to capture human activity data. As the number of electronic devices and their applications continues to grow, advances in artificial intelligence (AI) have revolutionized the ability to extract deeply hidden information for accurate recognition and interpretation. Human Activity Recognition (HAR) has received a lot of attention over the past two decades with applications such as remote health monitoring, security and surveillance, and smart environments. This study focuses on the recognition of human activity using smartphone sensors using different machine learning classification approaches. Data from accelerometer and gyroscope sensors in smartphones are used to recognize human activity. The aim of this study is to detect human movements by using the sensors of the smart phone. For this reason, sensor data for 18 different human movements were collected via smart phone and the generated data were tested with Logistic Regression, Decision Tree, Support Vector Classification and Random Forest machine learning methods and their performances were compared. The results show that Decision Tree and Logistic Regression can classify sub-samples with approximately 99% accuracy. This dataset will enable smartphones and other smart devices to identify new activities and help researchers develop finer models based on practical HAR data.