Human Activity Recognition Using Tri-Axial Angular Velocity

Abstract

Human Activity recognition is one of the prime applications of machine learning algorithms nowadays. It is being used in the field of healthcare, game development, smart environments, etc. Data from the sensors connected to a person can be employed to train supervised machine learning models in order to foresee the movement being done by the individual. In this paper, we will utilize data available at UCI machine learning Repository. It contains data generated from the accelerometer, gyroscope and other different sensors of Smartphone to train supervised predictive models using machine learning strategies like KNN, Logistic Regression, etc. to generate a model which can be used to predict the kind of movement being carried out by the individual which is branched into six activities like walking, walking upstairs, walking downstairs, sitting, standing and laying. We will be comparing the accuracy of our model using KNN and Logistic Regression.