Human Activity Recognition is an active field of research and is the basis for a wide range of modern day gadgets such
as fitness bands, sleep-tracking devices etc.[6] In this research based project, we present our efforts of executing a performance
comparison in order to determine the most effective machine learning algorithm for Human Activity Recognition applications.
Our objective in this project is to train Artificial Neural Networks, Random Forest Classifiers, k-NN and Support Vector
Machine algorithms over an HAR dataset (Human Activity Recognition using Smartphones Dataset) which is publicly available.
We then graphically compare the accuracy achieved by each of the algorithms thereby finding out the approach that should be
used by developers in Human Activity Recognition applications (such as fitness tracking platforms and home automation
platforms) in order to improvise and contrive their platforms.
