Estimating the extent of COVID19 Susceptibilities, Infections, and Recoveries using SIR-F model

Divyank Singh, Rakshit Goyal, Dr. Jatin Bedi
Thapar Institute of Engineering and Technology
July 2022
Paper Code

Abstract

The COVID-19 pandemic, which began in 2019 in Wuhan, spread globally at an alarming rate bringing manufacturing and trade activities worldwide to a standstill, thereby damaging the global economy and social well-being. Containing the spread of infection was given prime importance by all countries, leading to the demand for a model to predict the extent of the disease over a time period in the near future. In this paper, a SIR-F model predicting the extent of COVID-19 infections over the next 200 days is developed by studying the Susceptible Recovered ratio and observing trends in epidemiological parameters by using several datasets for three countries, namely the JHU dataset, OxCGRT dataset, PCRData and VaccineData. From the experimental analysis, it has been found that the Federal Republic of Germany and United States of America will be in the endemic by May-June 2022 while the Republic of India can witness a surge in number of infections, however, the intensity of infections will be significantly lower than previous instances. Upon plotting the obtained results against experimental data, the model was found to be accurate at estimating future trends.

The research aims at providing a deadline for the end of the COVID19 pandemic in 3 countries. In the proposed solution importance has been given to strigency measures such as vaccinations and lockdowns. Therefore, a compartmental model run with Ordinary Differential Equations was modelled by assigning the target population to three compartments- i) Susceptible, ii) Infected, and iii) Recovered. To minimise the Root Mean Square Logarithmic Error model hyperparameters were optimized. Conclusions were drawn after plotting the three population compartments and comparing the positivity rate and reproduction number.