Forecasting COVID-19 with gamma model

Abstract

COVID-19 is a highly contagiously atypical pneumonia attributed to a novel coronavirus. The global economy and people's lives have been tremendously affected by the COVID-19 pandemic since its outbreak in Wuhan, Hubei province, China. In this thesis, a non-linear model based on gamma distribution was built to verify the accuracy of the forecasting of the total confirmed cases of COVID-19 two weeks ahead. The daily growth in cases of COVID-19 for different countries was monitored and compared with the forecasted values. The verification of the performance of the non-linear Gamma distribution model has been verified by the non-linear regression. The data for the 19 countries with the most total confirmed COVID-19 cases as of June 22 was used. The data was sourced from the interactive web-based dashboard developed by the Center for System Science and Engineering (CSSE) at Johns Hopkins University. A web page has been developed to provide predictions generated by our models for individuals and public organizations to forecast the trends of COVID19.