Quantitative and Qualitative Analysis of Human Activities using Machine Learning Methods

Abstract

In the era of the digital world, every human activity needs to be in the count for various applications like health tracking, monitoring, and security so on. Smart devices like activity trackers, fit-bits, mobile phones, and smartwatches are helping a lot in activity monitoring. The insights of these monitoring devices greatly contributed to various research fields like activity recognition, classification, and gender and age classification etc. In this research, we used a standard dataset “Heterogeneity Activity Recognition Data Set” downloaded from UCI Machine learning Repository. Furthermore, quantitative and qualitative analysis carried out on human activity recognition (HAR) based on dataset attributes. In addition, we find the correlation between activities of each subject, and also activities recognized with 99% accuracy using machine learning techniques.