Data driven simulations of infectious diseases: Exploring facial recognition approach in predicting the infection severity

Abstract

Current approaches to the simulations of epidemiological diseases generally involve the classic compartmental spreading models such as SI, SIS and SIR, the use of which alongside graph theory provides a good understanding and prediction process for the simulations, and modeling of viral infections. Most models however set a transmission rate assuming homogeneity of the population, neglecting features such as age, gender or ethnicity, which have been found to impact the spread of the diseases and its severity. This paper presents a new experimental platform, which extends the current efforts of simulating epidemics, and considers the additional characteristics, in order to firstly be able to predict the infections more accurately, and secondly to allow to monitor the pandemic and assess infection risks, based on visual input.

Keywords

• Data science
• Facial recognition
• Compartmental modeling
• Simulations