Machine Learning based Digital Contact Tracing for Covid-19 in Myanmar

Abstract

Digital contact tracing is one of supportive tools for the prevention and reduction of wide spread of the covid-19 virus. Machine learning is one of the most supportive approaches for contact tracing applications. Thus, this research aims to develop a web-based contact tracing system by applying machine learning approaches. The purposes of the web-based contact tracing system are to spy out the contact cases, to support a short-term prediction for Covid-19 transmission rate, to give a warning or alert message to a person, who is in outbreak zone, to give a real-time information about Covid-19 to public in time, to collect everyday patient data in the same data standardized format from anywhere in Myanmar, and provide a richest dataset to other researchers via this website. The system is mainly composed of two sub-systems, one is digital contact tracing and other one is a short-term covid-19 transmission rate prediction. Digital contact tracing will give a prediction of possible hidden cases infected by a confirmed case by applying Neural Network algorithm and using confirmed- case-id, contact-date, contact-time-duration, contact-frequency, and contact-place.