Development of machine learning techniques for wearable vital signs monitoring device

Cardiovascular diseases (CVDs) are a wide-reaching prominent cause of death all over the world. According to the World Health Organization (WHO), approximately 17.9 million deaths are caused by CVDs each year. This accounts for 31% of all deaths that occurred worldwide. CVDs are related to heart problems like coronary artery disease, heart failure, and stroke. A low-cost, non-invasive method of measuring heart rate monitoring wearable is needed where heart rates can be intelligently monitored. The parameters that are used to measure the vital signs include Sinus tachycardia, Sinus bradycardia, normal sinus rhythm and stress levels. With the collected data, the project aims to develop machine learning techniques for cost effective wearable fitness trackers. As such, the project focuses on testing and collecting experimental data from an affordable fitness tracker. Most affordable trackers consist of optical heart rate sensors that offer low cost and high efficiency of battery life. It is essential to ensure that the raw data collected is accurate and consistent by comparing verified medical devices. Hence multiple tests need to be conducted to ensure the reliability of affordable fitness trackers to obtain reliable results. Through the test conducted Random Forest machine learning classifier has achieved the highest level of accuracy with the heart-rate data attribute and both the heart rate and stress level attributes. This model can be used to scale up with multiple attributes and easily applied to wearable fitness trackers, which would provide an affordable and convenient method of monitoring vital signs without the need for any sophisticated equipment.Bachelor of Engineering (Electrical and Electronic Engineering