Human activity recognition system using smartphone based on machine learning algorithms (original) (raw)

Human Activity Recognition System aims to capture the state of the user with respect to the external/heterogeneous environment in order to analyze the human health conditions with the help of the numerous sensors attached to the different body parts of the human 3. The first hand-held mobile was developed in the year 1979 and from that year to 2011 it is surveyed that around 80% of the world population is now using the Smartphone and from that year it is been observed that there is been a tremendous increase/up gradation in the field of technology and size also and in this futuristic world of technology. As we can see now a days Smartphone are playing a key role and are used to collect the data of human activities and which is further used to monitor the health of the person. Prior to collect the human activity data we use to attach the sensors to the different body parts like chest, hands, thighs, wrist etc. In this paper, with the help of the m ach in e learn in g algorithms and the data mining techniques we are going to analyze the posture by keeping in mind about the attributes which are going to take part. We are using six of the attributes with the help of these attributes we are going to predict the conditions which are as follows: Sitting, Standing, Walking, Walking Upstairs, Walking downstairs and Laying.