Development of an Intelligent Device capable of classifying Emotional Stress in correlation with Body Position

Abstract

The main aim of this work is to develop a device that is intelligent in terms of progressively learning conditional changes that occur in the human body during times of distress and taking action if necessary. This paper delves into the concept of Machine learning for the classification of a subject’s Emotional Stress Level and talks about the correlation of Emotional Stress with Activity Recognition. Additionally, it analyzes the significance of the correlation made. The paper deals with a device that has the capability of automatically perceiving any dangerous situation, and taking action if results prove to be positive i.e if subject is in danger. It does so by making use of an Open Source Machine Learning Toolkit- WEKA, which performs classification and evaluation of the data being recorded, thereby making the device intelligent. Analysis of all sensor data is done on MATLAB.